tests/testthat/test-spMsPGOcc-NNGP.R
In doserjef/spOccupancy: Single-Species, Multi-Species, and Integrated Spatial Occupancy Models
# Test spMsPGOcc.R -------------------------------------------------------
# NNGP --------------------------------------------------------------------
skip_on_cran()
# Intercept Only ----------------------------------------------------------
set.seed(1010)
J.x <- 8
J.y <- 8
J <- J.x * J.y
n.rep<- sample(2:4, size = J, replace = TRUE)
N <- 6
# Community-level covariate effects
# Occurrence
beta.mean <- c(0.2)
p.occ <- length(beta.mean)
tau.sq.beta <- c(0.6)
# Detection
alpha.mean <- c(0)
tau.sq.alpha <- c(1)
p.det <- length(alpha.mean)
# Random effects
psi.RE <- list()
p.RE <- list()
# Draw species-level effects from community means.
beta <- matrix(NA, nrow = N, ncol = p.occ)
alpha <- matrix(NA, nrow = N, ncol = p.det)
for (i in 1:p.occ) {
beta[, i] <- rnorm(N, beta.mean[i], sqrt(tau.sq.beta[i]))
}
for (i in 1:p.det) {
alpha[, i] <- rnorm(N, alpha.mean[i], sqrt(tau.sq.alpha[i]))
}
alpha.true <- alpha
phi <- rep(3 / .7, N)
sigma.sq <- rep(2, N)
nu <- rep(2, N)
dat <- simMsOcc(J.x = J.x, J.y = J.y, n.rep = n.rep, N = N, beta = beta, alpha = alpha,
psi.RE = psi.RE, p.RE = p.RE, sp = TRUE, sigma.sq = sigma.sq,
phi = phi, nu = nu, cov.model = 'matern')
pred.indx <- sample(1:J, round(J * .25), replace = FALSE)
y <- dat$y[, -pred.indx, , drop = FALSE]
# Occupancy covariates
X <- dat$X[-pred.indx, , drop = FALSE]
coords <- as.matrix(dat$coords[-pred.indx, , drop = FALSE])
# Prediction covariates
X.0 <- dat$X[pred.indx, , drop = FALSE]
coords.0 <- as.matrix(dat$coords[pred.indx, , drop = FALSE])
# Detection covariates
X.p <- dat$X.p[-pred.indx, , , drop = FALSE]
data.list <- list(y = y, coords = coords)
# Priors
prior.list <- list(beta.comm.normal = list(mean = 0, var = 2.72),
alpha.comm.normal = list(mean = 0, var = 2.72),
tau.sq.beta.ig = list(a = 0.1, b = 0.1),
tau.sq.alpha.ig = list(a = 0.1, b = 0.1),
nu.unif = list(0.5, 2.5))
# Starting values
inits.list <- list(alpha.comm = 0,
beta.comm = 0,
beta = 0,
alpha = 0,
tau.sq.beta = 1,
tau.sq.alpha = 1,
fix = TRUE,
z = apply(y, c(1, 2), max, na.rm = TRUE))
# Tuning
tuning.list <- list(phi = 1, nu = 0.25)
batch.length <- 25
n.batch <- 10
n.report <- 100
occ.formula <- ~ 1
det.formula <- ~ 1
out <- spMsPGOcc(occ.formula = occ.formula,
det.formula = det.formula,
data = data.list,
inits = inits.list,
n.batch = n.batch,
batch.length = batch.length,
accept.rate = 0.43,
priors = prior.list,
cov.model = "matern",
tuning = tuning.list,
n.omp.threads = 1,
verbose = FALSE,
NNGP = TRUE,
n.neighbors = 5,
search.type = 'cb',
n.report = 10,
n.burn = 100,
n.thin = 2,
n.chains = 2,
k.fold = 2,
k.fold.threads = 1)
# To make sure it worked --------------
test_that("out is of class spMsPGOcc", {
expect_s3_class(out, "spMsPGOcc")
})
# Check cross-validation --------------
test_that("cross-validation works", {
out.k.fold <- spMsPGOcc(occ.formula = occ.formula,
det.formula = det.formula,
data = data.list,
inits = inits.list,
n.batch = n.batch,
batch.length = batch.length,
accept.rate = 0.43,
priors = prior.list,
cov.model = "matern",
tuning = tuning.list,
n.omp.threads = 1,
verbose = FALSE,
NNGP = TRUE,
n.neighbors = 5,
search.type = 'cb',
n.report = 10,
n.burn = 100,
n.thin = 2,
n.chains = 2,
k.fold = 2,
k.fold.only = TRUE,
k.fold.threads = 1)
expect_equal(length(out.k.fold$k.fold.deviance), N)
expect_type(out.k.fold$k.fold.deviance, "double")
expect_equal(sum(out.k.fold$k.fold.deviance < 0), 0)
})
# Check random effects ----------------
test_that("random effects are correct", {
expect_equal(out$pRE, FALSE)
expect_equal(out$psiRE, FALSE)
})
# Check output data output is correct -
test_that("out$y == y", {
expect_equal(out$y, y)
})
# Check non-integer n.post -------------
test_that("non-integer n.post", {
expect_error(out <- spMsPGOcc(occ.formula = occ.formula,
det.formula = det.formula,
data = data.list,
n.thin = 13,
n.batch = n.batch,
batch.length = batch.length,
accept.rate = 0.43,
n.omp.threads = 1,
verbose = FALSE))
})
# Check default values ----------------
test_that("default priors, inits, burn, thin work", {
out <- spMsPGOcc(occ.formula = occ.formula,
det.formula = det.formula,
data = data.list,
n.batch = n.batch,
batch.length = batch.length,
accept.rate = 0.43,
cov.model = "exponential",
tuning = tuning.list,
NNGP = TRUE,
verbose = FALSE,
n.neighbors = 5,
n.chains = 1)
expect_s3_class(out, "spMsPGOcc")
})
test_that("all correlation functions work", {
out <- spMsPGOcc(occ.formula = occ.formula,
det.formula = det.formula,
data = data.list,
inits = inits.list,
n.batch = n.batch,
batch.length = batch.length,
accept.rate = 0.43,
priors = prior.list,
cov.model = "gaussian",
tuning = list(phi = 0.3),
n.omp.threads = 1,
verbose = FALSE,
NNGP = TRUE,
n.neighbors = 5,
search.type = 'cb',
n.report = 10,
n.burn = 100,
n.thin = 1,
n.chains = 1)
expect_s3_class(out, "spMsPGOcc")
out <- spMsPGOcc(occ.formula = occ.formula,
det.formula = det.formula,
data = data.list,
inits = inits.list,
n.batch = n.batch,
batch.length = batch.length,
accept.rate = 0.43,
priors = prior.list,
cov.model = "spherical",
tuning = list(phi = 0.3),
n.omp.threads = 1,
verbose = FALSE,
NNGP = TRUE,
n.neighbors = 5,
search.type = 'cb',
n.report = 10,
n.burn = 100,
n.thin = 1,
n.chains = 1)
expect_s3_class(out, "spMsPGOcc")
out <- spMsPGOcc(occ.formula = occ.formula,
det.formula = det.formula,
data = data.list,
inits = inits.list,
n.batch = n.batch,
batch.length = batch.length,
accept.rate = 0.43,
priors = list(nu.unif = list(0.4, 3)),
cov.model = "matern",
tuning = list(phi = 0.3, nu = 0.2),
n.omp.threads = 1,
verbose = FALSE,
NNGP = TRUE,
n.neighbors = 5,
search.type = 'cb',
n.report = 10,
n.burn = 100,
n.thin = 1,
n.chains = 1)
expect_s3_class(out, "spMsPGOcc")
})
test_that("verbose prints to the screen", {
expect_output(spMsPGOcc(occ.formula = occ.formula,
det.formula = det.formula,
data = data.list,
inits = inits.list,
n.batch = n.batch,
batch.length = batch.length,
accept.rate = 0.43,
priors = prior.list,
cov.model = "exponential",
tuning = tuning.list,
n.omp.threads = 1,
verbose = TRUE,
NNGP = TRUE,
n.neighbors = 5,
search.type = 'cb',
n.report = 10,
n.burn = 100,
n.thin = 1,
n.chains = 1))
})
# Check waicOcc -----------------------
test_that("waicOCC works for spMsPGOcc", {
# as.vector gets rid of names
waic.out <- as.vector(waicOcc(out))
expect_equal(length(waic.out), 3)
expect_equal(waic.out[3], -2 * (waic.out[1] - waic.out[2]))
})
test_that("fitted works for spMsPGOcc", {
fitted.out <- fitted(out)
expect_equal(length(fitted.out), 2)
expect_equal(class(fitted.out$y.rep.samples), "array")
expect_equal(class(fitted.out$p.samples), "array")
expect_equal(dim(fitted.out$y.rep.samples), dim(fitted.out$p.samples))
})
test_that("predict works for spMsPGOcc", {
n.post.samples <- out$n.post * out$n.chains
pred.out <- predict(out, X.0, coords.0, verbose = FALSE)
expect_type(pred.out, "list")
expect_equal(dim(pred.out$psi.0.samples), c(n.post.samples, N, nrow(X.0)))
expect_equal(dim(pred.out$z.0.samples), c(n.post.samples, N, nrow(X.0)))
})
test_that("detection prediction works", {
J.str <- 100
X.p.0 <- matrix(1, nrow = J.str, ncol = p.det)
pred.out <- predict(out, X.p.0, type = 'detection')
expect_type(pred.out, 'list')
expect_equal(dim(pred.out$p.0.samples), c(out$n.post * out$n.chains, N, J.str))
})
test_that("posterior predictive checks work for msPGOcc", {
n.post.samples <- out$n.post * out$n.chains
J.fit <- nrow(X)
ppc.out <- ppcOcc(out, 'chi-square', 2)
expect_type(ppc.out, "list")
expect_equal(dim(ppc.out$fit.y), c(n.post.samples, N))
expect_equal(dim(ppc.out$fit.y.rep), c(n.post.samples, N))
expect_equal(dim(ppc.out$fit.y.group.quants), c(5, N, max(n.rep)))
expect_equal(dim(ppc.out$fit.y.rep.group.quants), c(5, N, max(n.rep)))
ppc.out <- ppcOcc(out, 'chi-square', 1)
expect_type(ppc.out, "list")
expect_equal(dim(ppc.out$fit.y), c(n.post.samples, N))
expect_equal(dim(ppc.out$fit.y.rep), c(n.post.samples, N))
expect_equal(dim(ppc.out$fit.y.group.quants), c(5, N, J.fit))
expect_equal(dim(ppc.out$fit.y.rep.group.quants), c(5, N, J.fit))
ppc.out <- ppcOcc(out, 'freeman-tukey', 1)
expect_type(ppc.out, "list")
expect_equal(dim(ppc.out$fit.y), c(n.post.samples, N))
expect_equal(dim(ppc.out$fit.y.rep), c(n.post.samples, N))
expect_equal(dim(ppc.out$fit.y.group.quants), c(5, N, J.fit))
expect_equal(dim(ppc.out$fit.y.rep.group.quants), c(5, N, J.fit))
ppc.out <- ppcOcc(out, 'freeman-tukey', 2)
expect_type(ppc.out, "list")
expect_equal(dim(ppc.out$fit.y), c(n.post.samples, N))
expect_equal(dim(ppc.out$fit.y.rep), c(n.post.samples, N))
expect_equal(dim(ppc.out$fit.y.group.quants), c(5, N, max(n.rep)))
expect_equal(dim(ppc.out$fit.y.rep.group.quants), c(5, N, max(n.rep)))
})
# Occurrence coviarate only -----------------------------------------------
J.x <- 8
J.y <- 8
J <- J.x * J.y
n.rep<- sample(2:4, size = J, replace = TRUE)
N <- 6
# Community-level covariate effects
# Occurrence
beta.mean <- c(0.2, 0.5)
p.occ <- length(beta.mean)
tau.sq.beta <- c(0.6, 2.3)
# Detection
alpha.mean <- c(0)
tau.sq.alpha <- c(1)
p.det <- length(alpha.mean)
# Random effects
psi.RE <- list()
p.RE <- list()
# Draw species-level effects from community means.
beta <- matrix(NA, nrow = N, ncol = p.occ)
alpha <- matrix(NA, nrow = N, ncol = p.det)
for (i in 1:p.occ) {
beta[, i] <- rnorm(N, beta.mean[i], sqrt(tau.sq.beta[i]))
}
for (i in 1:p.det) {
alpha[, i] <- rnorm(N, alpha.mean[i], sqrt(tau.sq.alpha[i]))
}
alpha.true <- alpha
phi <- rep(3 / .7, N)
sigma.sq <- rep(2, N)
nu <- rep(2, N)
dat <- simMsOcc(J.x = J.x, J.y = J.y, n.rep = n.rep, N = N, beta = beta, alpha = alpha,
psi.RE = psi.RE, p.RE = p.RE, sp = TRUE, sigma.sq = sigma.sq,
phi = phi, nu = nu, cov.model = 'matern')
pred.indx <- sample(1:J, round(J * .25), replace = FALSE)
y <- dat$y[, -pred.indx, , drop = FALSE]
# Occupancy covariates
X <- dat$X[-pred.indx, , drop = FALSE]
coords <- as.matrix(dat$coords[-pred.indx, , drop = FALSE])
# Prediction covariates
X.0 <- dat$X[pred.indx, , drop = FALSE]
coords.0 <- as.matrix(dat$coords[pred.indx, , drop = FALSE])
# Detection covariates
X.p <- dat$X.p[-pred.indx, , , drop = FALSE]
occ.covs <- X
colnames(occ.covs) <- c('int', 'occ.cov.1')
data.list <- list(y = y, coords = coords, occ.covs = occ.covs)
# Priors
prior.list <- list(beta.comm.normal = list(mean = 0, var = 2.72),
alpha.comm.normal = list(mean = 0, var = 2.72),
tau.sq.beta.ig = list(a = 0.1, b = 0.1),
tau.sq.alpha.ig = list(a = 0.1, b = 0.1),
nu.unif = list(0.5, 2.5))
# Starting values
inits.list <- list(alpha.comm = 0,
beta.comm = 0,
beta = 0,
alpha = 0,
tau.sq.beta = 1,
tau.sq.alpha = 1,
z = apply(y, c(1, 2), max, na.rm = TRUE))
# Tuning
tuning.list <- list(phi = 1, nu = 0.25)
batch.length <- 25
n.batch <- 10
n.report <- 100
occ.formula <- ~ occ.cov.1
det.formula <- ~ 1
out <- spMsPGOcc(occ.formula = occ.formula,
det.formula = det.formula,
data = data.list,
inits = inits.list,
n.batch = n.batch,
batch.length = batch.length,
accept.rate = 0.43,
priors = prior.list,
cov.model = "matern",
tuning = tuning.list,
n.omp.threads = 1,
verbose = FALSE,
NNGP = TRUE,
n.neighbors = 5,
search.type = 'cb',
n.report = 10,
n.burn = 100,
n.thin = 2,
n.chains = 2,
k.fold = 2,
k.fold.threads = 1)
# To make sure it worked --------------
test_that("out is of class spMsPGOcc", {
expect_s3_class(out, "spMsPGOcc")
})
# Check cross-validation --------------
test_that("cross-validation works", {
expect_equal(length(out$k.fold.deviance), N)
expect_type(out$k.fold.deviance, "double")
expect_equal(sum(out$k.fold.deviance < 0), 0)
})
# Check random effects ----------------
test_that("random effects are correct", {
expect_equal(out$pRE, FALSE)
expect_equal(out$psiRE, FALSE)
})
# Check output data output is correct -
test_that("out$y == y", {
expect_equal(out$y, y)
})
# Check missing values ----------------
test_that("missing value error handling works", {
tmp.data <- data.list
tmp.data$occ.covs[3, ] <- NA
expect_error(spMsPGOcc(occ.formula = occ.formula,
det.formula = det.formula,
data = tmp.data,
n.batch = 40,
batch.length = batch.length,
cov.model = "exponential",
tuning = tuning.list,
NNGP = TRUE,
verbose = FALSE,
n.neighbors = 5,
search.type = 'cb',
n.report = 10,
n.burn = 100,
n.chains = 1))
# tmp.data <- data.list
# tmp.data$det.covs[[1]][1] <- NA
# expect_error(spMsPGOcc(occ.formula = occ.formula,
# det.formula = det.formula,
# data = tmp.data,
# n.batch = 40,
# batch.length = batch.length,
# cov.model = "exponential",
# tuning = tuning.list,
# NNGP = TRUE,
# verbose = FALSE,
# n.neighbors = 5,
# search.type = 'cb',
# n.report = 10,
# n.burn = 100,
# n.chains = 1))
# tmp.data <- data.list
# tmp.data$y[, 1, 1] <- NA
# out <- spMsPGOcc(occ.formula = occ.formula,
# det.formula = det.formula,
# data = tmp.data,
# n.batch = 40,
# batch.length = batch.length,
# cov.model = "exponential",
# tuning = tuning.list,
# NNGP = TRUE,
# verbose = FALSE,
# n.neighbors = 5,
# search.type = 'cb',
# n.report = 10,
# n.burn = 100,
# n.chains = 1)
# expect_s3_class(out, "spMsPGOcc")
})
# Check default values ----------------
test_that("default priors, inits, burn, thin work", {
out <- spMsPGOcc(occ.formula = occ.formula,
det.formula = det.formula,
data = data.list,
n.batch = n.batch,
batch.length = batch.length,
accept.rate = 0.43,
cov.model = "exponential",
tuning = tuning.list,
NNGP = TRUE,
verbose = FALSE,
n.neighbors = 5,
n.chains = 1)
expect_s3_class(out, "spMsPGOcc")
})
test_that("all correlation functions work", {
out <- spMsPGOcc(occ.formula = occ.formula,
det.formula = det.formula,
data = data.list,
inits = inits.list,
n.batch = n.batch,
batch.length = batch.length,
accept.rate = 0.43,
priors = prior.list,
cov.model = "gaussian",
tuning = list(phi = 0.3),
n.omp.threads = 1,
verbose = FALSE,
NNGP = TRUE,
n.neighbors = 5,
search.type = 'cb',
n.report = 10,
n.burn = 100,
n.thin = 1,
n.chains = 1)
expect_s3_class(out, "spMsPGOcc")
out <- spMsPGOcc(occ.formula = occ.formula,
det.formula = det.formula,
data = data.list,
inits = inits.list,
n.batch = n.batch,
batch.length = batch.length,
accept.rate = 0.43,
priors = prior.list,
cov.model = "spherical",
tuning = list(phi = 0.3),
n.omp.threads = 1,
verbose = FALSE,
NNGP = TRUE,
n.neighbors = 5,
search.type = 'cb',
n.report = 10,
n.burn = 100,
n.thin = 1,
n.chains = 1)
expect_s3_class(out, "spMsPGOcc")
out <- spMsPGOcc(occ.formula = occ.formula,
det.formula = det.formula,
data = data.list,
inits = inits.list,
n.batch = n.batch,
batch.length = batch.length,
accept.rate = 0.43,
priors = list(nu.unif = list(0.4, 3)),
cov.model = "matern",
tuning = list(phi = 0.3, nu = 0.2),
n.omp.threads = 1,
verbose = FALSE,
NNGP = TRUE,
n.neighbors = 5,
search.type = 'cb',
n.report = 10,
n.burn = 100,
n.thin = 1,
n.chains = 1)
expect_s3_class(out, "spMsPGOcc")
})
test_that("verbose prints to the screen", {
expect_output(spMsPGOcc(occ.formula = occ.formula,
det.formula = det.formula,
data = data.list,
inits = inits.list,
n.batch = n.batch,
batch.length = batch.length,
accept.rate = 0.43,
priors = prior.list,
cov.model = "exponential",
tuning = tuning.list,
n.omp.threads = 1,
verbose = TRUE,
NNGP = TRUE,
n.neighbors = 5,
search.type = 'cb',
n.report = 10,
n.burn = 100,
n.thin = 1,
n.chains = 1))
})
# Check waicOcc -----------------------
test_that("waicOCC works for spMsPGOcc", {
# as.vector gets rid of names
waic.out <- as.vector(waicOcc(out))
expect_equal(length(waic.out), 3)
expect_equal(waic.out[3], -2 * (waic.out[1] - waic.out[2]))
})
test_that("fitted works for spMsPGOcc", {
fitted.out <- fitted(out)
expect_equal(length(fitted.out), 2)
expect_equal(class(fitted.out$y.rep.samples), "array")
expect_equal(class(fitted.out$p.samples), "array")
expect_equal(dim(fitted.out$y.rep.samples), dim(fitted.out$p.samples))
})
test_that("predict works for spMsPGOcc", {
n.post.samples <- out$n.post * out$n.chains
pred.out <- predict(out, X.0, coords.0, verbose = FALSE)
expect_type(pred.out, "list")
expect_equal(dim(pred.out$psi.0.samples), c(n.post.samples, N, nrow(X.0)))
expect_equal(dim(pred.out$z.0.samples), c(n.post.samples, N, nrow(X.0)))
})
test_that("detection prediction works", {
J.str <- 100
X.p.0 <- matrix(1, nrow = J.str, ncol = p.det)
pred.out <- predict(out, X.p.0, type = 'detection')
expect_type(pred.out, 'list')
expect_equal(dim(pred.out$p.0.samples), c(out$n.post * out$n.chains, N, J.str))
})
test_that("posterior predictive checks work for msPGOcc", {
n.post.samples <- out$n.post * out$n.chains
J.fit <- nrow(X)
ppc.out <- ppcOcc(out, 'chi-square', 2)
expect_type(ppc.out, "list")
expect_equal(dim(ppc.out$fit.y), c(n.post.samples, N))
expect_equal(dim(ppc.out$fit.y.rep), c(n.post.samples, N))
expect_equal(dim(ppc.out$fit.y.group.quants), c(5, N, max(n.rep)))
expect_equal(dim(ppc.out$fit.y.rep.group.quants), c(5, N, max(n.rep)))
ppc.out <- ppcOcc(out, 'chi-square', 1)
expect_type(ppc.out, "list")
expect_equal(dim(ppc.out$fit.y), c(n.post.samples, N))
expect_equal(dim(ppc.out$fit.y.rep), c(n.post.samples, N))
expect_equal(dim(ppc.out$fit.y.group.quants), c(5, N, J.fit))
expect_equal(dim(ppc.out$fit.y.rep.group.quants), c(5, N, J.fit))
ppc.out <- ppcOcc(out, 'freeman-tukey', 1)
expect_type(ppc.out, "list")
expect_equal(dim(ppc.out$fit.y), c(n.post.samples, N))
expect_equal(dim(ppc.out$fit.y.rep), c(n.post.samples, N))
expect_equal(dim(ppc.out$fit.y.group.quants), c(5, N, J.fit))
expect_equal(dim(ppc.out$fit.y.rep.group.quants), c(5, N, J.fit))
ppc.out <- ppcOcc(out, 'freeman-tukey', 2)
expect_type(ppc.out, "list")
expect_equal(dim(ppc.out$fit.y), c(n.post.samples, N))
expect_equal(dim(ppc.out$fit.y.rep), c(n.post.samples, N))
expect_equal(dim(ppc.out$fit.y.group.quants), c(5, N, max(n.rep)))
expect_equal(dim(ppc.out$fit.y.rep.group.quants), c(5, N, max(n.rep)))
})
# Detection covariate only ------------------------------------------------
J.x <- 8
J.y <- 8
J <- J.x * J.y
n.rep<- sample(2:4, size = J, replace = TRUE)
N <- 6
# Community-level covariate effects
# Occurrence
beta.mean <- c(0.2)
p.occ <- length(beta.mean)
tau.sq.beta <- c(0.6)
# Detection
alpha.mean <- c(0, -0.5)
tau.sq.alpha <- c(1, 2.3)
p.det <- length(alpha.mean)
# Random effects
psi.RE <- list()
p.RE <- list()
# Draw species-level effects from community means.
beta <- matrix(NA, nrow = N, ncol = p.occ)
alpha <- matrix(NA, nrow = N, ncol = p.det)
for (i in 1:p.occ) {
beta[, i] <- rnorm(N, beta.mean[i], sqrt(tau.sq.beta[i]))
}
for (i in 1:p.det) {
alpha[, i] <- rnorm(N, alpha.mean[i], sqrt(tau.sq.alpha[i]))
}
alpha.true <- alpha
phi <- rep(3 / .7, N)
sigma.sq <- rep(2, N)
nu <- rep(2, N)
dat <- simMsOcc(J.x = J.x, J.y = J.y, n.rep = n.rep, N = N, beta = beta, alpha = alpha,
psi.RE = psi.RE, p.RE = p.RE, sp = TRUE, sigma.sq = sigma.sq,
phi = phi, nu = nu, cov.model = 'matern')
pred.indx <- sample(1:J, round(J * .25), replace = FALSE)
y <- dat$y[, -pred.indx, , drop = FALSE]
# Occupancy covariates
X <- dat$X[-pred.indx, , drop = FALSE]
coords <- as.matrix(dat$coords[-pred.indx, , drop = FALSE])
# Prediction covariates
X.0 <- dat$X[pred.indx, , drop = FALSE]
coords.0 <- as.matrix(dat$coords[pred.indx, , drop = FALSE])
# Detection covariates
X.p <- dat$X.p[-pred.indx, , , drop = FALSE]
occ.covs <- X
colnames(occ.covs) <- c('int')
det.covs <- list(det.cov.1 = X.p[, , 2])
data.list <- list(y = y, coords = coords, occ.covs = occ.covs, det.covs = det.covs)
# Priors
prior.list <- list(beta.comm.normal = list(mean = 0, var = 2.72),
alpha.comm.normal = list(mean = 0, var = 2.72),
tau.sq.beta.ig = list(a = 0.1, b = 0.1),
tau.sq.alpha.ig = list(a = 0.1, b = 0.1),
nu.unif = list(0.5, 2.5))
# Starting values
inits.list <- list(alpha.comm = 0,
beta.comm = 0,
beta = 0,
alpha = 0,
tau.sq.beta = 1,
tau.sq.alpha = 1,
z = apply(y, c(1, 2), max, na.rm = TRUE))
# Tuning
tuning.list <- list(phi = 1, nu = 0.25)
batch.length <- 25
n.batch <- 10
n.report <- 100
occ.formula <- ~ 1
det.formula <- ~ det.cov.1
out <- spMsPGOcc(occ.formula = occ.formula,
det.formula = det.formula,
data = data.list,
inits = inits.list,
n.batch = n.batch,
batch.length = batch.length,
accept.rate = 0.43,
priors = prior.list,
cov.model = "matern",
tuning = tuning.list,
n.omp.threads = 1,
verbose = FALSE,
NNGP = TRUE,
n.neighbors = 5,
search.type = 'cb',
n.report = 10,
n.burn = 100,
n.thin = 2,
n.chains = 2,
k.fold = 2,
k.fold.threads = 1)
# To make sure it worked --------------
test_that("out is of class spMsPGOcc", {
expect_s3_class(out, "spMsPGOcc")
})
# Check cross-validation --------------
test_that("cross-validation works", {
expect_equal(length(out$k.fold.deviance), N)
expect_type(out$k.fold.deviance, "double")
expect_equal(sum(out$k.fold.deviance < 0), 0)
})
# Check random effects ----------------
test_that("random effects are correct", {
expect_equal(out$pRE, FALSE)
expect_equal(out$psiRE, FALSE)
})
# Check output data output is correct -
test_that("out$y == y", {
expect_equal(out$y, y)
})
# Check missing values ----------------
test_that("missing value error handling works", {
# tmp.data <- data.list
# tmp.data$occ.covs[3, ] <- NA
# expect_error(spMsPGOcc(occ.formula = occ.formula,
# det.formula = det.formula,
# data = tmp.data,
# n.batch = 40,
# batch.length = batch.length,
# cov.model = "exponential",
# tuning = tuning.list,
# NNGP = TRUE,
# verbose = FALSE,
# n.neighbors = 5,
# search.type = 'cb',
# n.report = 10,
# n.burn = 100,
# n.chains = 1))
tmp.data <- data.list
tmp.data$y[1, 1, 1] <- 1
tmp.data$det.covs[[1]][1] <- NA
expect_error(spMsPGOcc(occ.formula = occ.formula,
det.formula = det.formula,
data = tmp.data,
n.batch = 40,
batch.length = batch.length,
cov.model = "exponential",
tuning = tuning.list,
NNGP = TRUE,
verbose = FALSE,
n.neighbors = 5,
search.type = 'cb',
n.report = 10,
n.burn = 100,
n.chains = 1))
tmp.data <- data.list
tmp.data$y[, 1, 1] <- NA
out <- spMsPGOcc(occ.formula = occ.formula,
det.formula = det.formula,
data = tmp.data,
n.batch = 40,
batch.length = batch.length,
cov.model = "exponential",
tuning = tuning.list,
NNGP = TRUE,
verbose = FALSE,
n.neighbors = 5,
search.type = 'cb',
n.report = 10,
n.burn = 100,
n.chains = 1)
expect_s3_class(out, "spMsPGOcc")
})
# Check default values ----------------
test_that("default priors, inits, burn, thin work", {
out <- spMsPGOcc(occ.formula = occ.formula,
det.formula = det.formula,
data = data.list,
n.batch = n.batch,
batch.length = batch.length,
accept.rate = 0.43,
cov.model = "exponential",
tuning = tuning.list,
NNGP = TRUE,
verbose = FALSE,
n.neighbors = 5,
n.chains = 1)
expect_s3_class(out, "spMsPGOcc")
})
test_that("all correlation functions work", {
out <- spMsPGOcc(occ.formula = occ.formula,
det.formula = det.formula,
data = data.list,
inits = inits.list,
n.batch = n.batch,
batch.length = batch.length,
accept.rate = 0.43,
priors = prior.list,
cov.model = "gaussian",
tuning = list(phi = 0.3),
n.omp.threads = 1,
verbose = FALSE,
NNGP = TRUE,
n.neighbors = 5,
search.type = 'cb',
n.report = 10,
n.burn = 100,
n.thin = 1,
n.chains = 1)
expect_s3_class(out, "spMsPGOcc")
out <- spMsPGOcc(occ.formula = occ.formula,
det.formula = det.formula,
data = data.list,
inits = inits.list,
n.batch = n.batch,
batch.length = batch.length,
accept.rate = 0.43,
priors = prior.list,
cov.model = "spherical",
tuning = list(phi = 0.3),
n.omp.threads = 1,
verbose = FALSE,
NNGP = TRUE,
n.neighbors = 5,
search.type = 'cb',
n.report = 10,
n.burn = 100,
n.thin = 1,
n.chains = 1)
expect_s3_class(out, "spMsPGOcc")
out <- spMsPGOcc(occ.formula = occ.formula,
det.formula = det.formula,
data = data.list,
inits = inits.list,
n.batch = n.batch,
batch.length = batch.length,
accept.rate = 0.43,
priors = list(nu.unif = list(0.4, 3)),
cov.model = "matern",
tuning = list(phi = 0.3, nu = 0.2),
n.omp.threads = 1,
verbose = FALSE,
NNGP = TRUE,
n.neighbors = 5,
search.type = 'cb',
n.report = 10,
n.burn = 100,
n.thin = 1,
n.chains = 1)
expect_s3_class(out, "spMsPGOcc")
})
test_that("verbose prints to the screen", {
expect_output(spMsPGOcc(occ.formula = occ.formula,
det.formula = det.formula,
data = data.list,
inits = inits.list,
n.batch = n.batch,
batch.length = batch.length,
accept.rate = 0.43,
priors = prior.list,
cov.model = "exponential",
tuning = tuning.list,
n.omp.threads = 1,
verbose = TRUE,
NNGP = TRUE,
n.neighbors = 5,
search.type = 'cb',
n.report = 10,
n.burn = 100,
n.thin = 1,
n.chains = 1))
})
# Check waicOcc -----------------------
test_that("waicOCC works for spMsPGOcc", {
# as.vector gets rid of names
waic.out <- as.vector(waicOcc(out))
expect_equal(length(waic.out), 3)
expect_equal(waic.out[3], -2 * (waic.out[1] - waic.out[2]))
})
test_that("fitted works for spMsPGOcc", {
fitted.out <- fitted(out)
expect_equal(length(fitted.out), 2)
expect_equal(class(fitted.out$y.rep.samples), "array")
expect_equal(class(fitted.out$p.samples), "array")
expect_equal(dim(fitted.out$y.rep.samples), dim(fitted.out$p.samples))
})
test_that("predict works for spMsPGOcc", {
n.post.samples <- out$n.post * out$n.chains
pred.out <- predict(out, X.0, coords.0, verbose = FALSE)
expect_type(pred.out, "list")
expect_equal(dim(pred.out$psi.0.samples), c(n.post.samples, N, nrow(X.0)))
expect_equal(dim(pred.out$z.0.samples), c(n.post.samples, N, nrow(X.0)))
})
test_that("detection prediction works", {
X.p.0 <- dat$X.p[, 1, ]
pred.out <- predict(out, X.p.0, type = 'detection')
expect_type(pred.out, 'list')
expect_equal(dim(pred.out$p.0.samples), c(out$n.post * out$n.chains, N, J))
})
test_that("posterior predictive checks work for msPGOcc", {
n.post.samples <- out$n.post * out$n.chains
J.fit <- nrow(X)
ppc.out <- ppcOcc(out, 'chi-square', 2)
expect_type(ppc.out, "list")
expect_equal(dim(ppc.out$fit.y), c(n.post.samples, N))
expect_equal(dim(ppc.out$fit.y.rep), c(n.post.samples, N))
expect_equal(dim(ppc.out$fit.y.group.quants), c(5, N, max(n.rep)))
expect_equal(dim(ppc.out$fit.y.rep.group.quants), c(5, N, max(n.rep)))
ppc.out <- ppcOcc(out, 'chi-square', 1)
expect_type(ppc.out, "list")
expect_equal(dim(ppc.out$fit.y), c(n.post.samples, N))
expect_equal(dim(ppc.out$fit.y.rep), c(n.post.samples, N))
expect_equal(dim(ppc.out$fit.y.group.quants), c(5, N, J.fit))
expect_equal(dim(ppc.out$fit.y.rep.group.quants), c(5, N, J.fit))
ppc.out <- ppcOcc(out, 'freeman-tukey', 1)
expect_type(ppc.out, "list")
expect_equal(dim(ppc.out$fit.y), c(n.post.samples, N))
expect_equal(dim(ppc.out$fit.y.rep), c(n.post.samples, N))
expect_equal(dim(ppc.out$fit.y.group.quants), c(5, N, J.fit))
expect_equal(dim(ppc.out$fit.y.rep.group.quants), c(5, N, J.fit))
ppc.out <- ppcOcc(out, 'freeman-tukey', 2)
expect_type(ppc.out, "list")
expect_equal(dim(ppc.out$fit.y), c(n.post.samples, N))
expect_equal(dim(ppc.out$fit.y.rep), c(n.post.samples, N))
expect_equal(dim(ppc.out$fit.y.group.quants), c(5, N, max(n.rep)))
expect_equal(dim(ppc.out$fit.y.rep.group.quants), c(5, N, max(n.rep)))
})
# Covariates on both ------------------------------------------------------
J.x <- 8
J.y <- 8
J <- J.x * J.y
n.rep<- sample(2:4, size = J, replace = TRUE)
N <- 6
# Community-level covariate effects
# Occurrence
beta.mean <- c(0.2, 0.5, 1.2)
p.occ <- length(beta.mean)
tau.sq.beta <- c(0.6, 1.5, 2.3)
# Detection
alpha.mean <- c(0, -0.5)
tau.sq.alpha <- c(1, 2.3)
p.det <- length(alpha.mean)
# Random effects
psi.RE <- list()
p.RE <- list()
# Draw species-level effects from community means.
beta <- matrix(NA, nrow = N, ncol = p.occ)
alpha <- matrix(NA, nrow = N, ncol = p.det)
for (i in 1:p.occ) {
beta[, i] <- rnorm(N, beta.mean[i], sqrt(tau.sq.beta[i]))
}
for (i in 1:p.det) {
alpha[, i] <- rnorm(N, alpha.mean[i], sqrt(tau.sq.alpha[i]))
}
alpha.true <- alpha
phi <- rep(3 / .7, N)
sigma.sq <- rep(2, N)
nu <- rep(2, N)
dat <- simMsOcc(J.x = J.x, J.y = J.y, n.rep = n.rep, N = N, beta = beta, alpha = alpha,
psi.RE = psi.RE, p.RE = p.RE, sp = TRUE, sigma.sq = sigma.sq,
phi = phi, nu = nu, cov.model = 'matern')
pred.indx <- sample(1:J, round(J * .25), replace = FALSE)
y <- dat$y[, -pred.indx, , drop = FALSE]
# Occupancy covariates
X <- dat$X[-pred.indx, , drop = FALSE]
coords <- as.matrix(dat$coords[-pred.indx, , drop = FALSE])
# Prediction covariates
X.0 <- dat$X[pred.indx, , drop = FALSE]
coords.0 <- as.matrix(dat$coords[pred.indx, , drop = FALSE])
# Detection covariates
X.p <- dat$X.p[-pred.indx, , , drop = FALSE]
occ.covs <- X
colnames(occ.covs) <- c('int', 'occ.cov.1', 'occ.cov.2')
det.covs <- list(det.cov.1 = X.p[, , 2])
data.list <- list(y = y, coords = coords, occ.covs = occ.covs, det.covs = det.covs)
# Priors
prior.list <- list(beta.comm.normal = list(mean = 0, var = 2.72),
alpha.comm.normal = list(mean = 0, var = 2.72),
tau.sq.beta.ig = list(a = 0.1, b = 0.1),
tau.sq.alpha.ig = list(a = 0.1, b = 0.1),
nu.unif = list(0.5, 2.5))
# Starting values
inits.list <- list(alpha.comm = 0,
beta.comm = 0,
beta = 0,
alpha = 0,
tau.sq.beta = 1,
tau.sq.alpha = 1,
z = apply(y, c(1, 2), max, na.rm = TRUE))
# Tuning
tuning.list <- list(phi = 1, nu = 0.25)
batch.length <- 25
n.batch <- 10
n.report <- 100
occ.formula <- ~ occ.cov.1 + occ.cov.2
det.formula <- ~ det.cov.1
out <- spMsPGOcc(occ.formula = occ.formula,
det.formula = det.formula,
data = data.list,
inits = inits.list,
n.batch = n.batch,
batch.length = batch.length,
accept.rate = 0.43,
priors = prior.list,
cov.model = "matern",
tuning = tuning.list,
n.omp.threads = 1,
verbose = FALSE,
NNGP = TRUE,
n.neighbors = 5,
search.type = 'cb',
n.report = 10,
n.burn = 100,
n.thin = 2,
n.chains = 2,
k.fold = 2,
k.fold.threads = 1)
# To make sure it worked --------------
test_that("out is of class spMsPGOcc", {
expect_s3_class(out, "spMsPGOcc")
})
# Check cross-validation --------------
test_that("cross-validation works", {
expect_equal(length(out$k.fold.deviance), N)
expect_type(out$k.fold.deviance, "double")
expect_equal(sum(out$k.fold.deviance < 0), 0)
})
# Check random effects ----------------
test_that("random effects are correct", {
expect_equal(out$pRE, FALSE)
expect_equal(out$psiRE, FALSE)
})
# Check output data output is correct -
test_that("out$y == y", {
expect_equal(out$y, y)
})
# Check missing values ----------------
test_that("missing value error handling works", {
tmp.data <- data.list
tmp.data$occ.covs[3, ] <- NA
expect_error(spMsPGOcc(occ.formula = occ.formula,
det.formula = det.formula,
data = tmp.data,
n.batch = 40,
batch.length = batch.length,
cov.model = "exponential",
tuning = tuning.list,
NNGP = TRUE,
verbose = FALSE,
n.neighbors = 5,
search.type = 'cb',
n.report = 10,
n.burn = 100,
n.chains = 1))
tmp.data <- data.list
tmp.data$y[1, 1, 1] <- 1
tmp.data$det.covs[[1]][1] <- NA
expect_error(spMsPGOcc(occ.formula = occ.formula,
det.formula = det.formula,
data = tmp.data,
n.batch = 40,
batch.length = batch.length,
cov.model = "exponential",
tuning = tuning.list,
NNGP = TRUE,
verbose = FALSE,
n.neighbors = 5,
search.type = 'cb',
n.report = 10,
n.burn = 100,
n.chains = 1))
tmp.data <- data.list
tmp.data$y[, 1, 1] <- NA
out <- spMsPGOcc(occ.formula = occ.formula,
det.formula = det.formula,
data = tmp.data,
n.batch = 40,
batch.length = batch.length,
cov.model = "exponential",
tuning = tuning.list,
NNGP = TRUE,
verbose = FALSE,
n.neighbors = 5,
search.type = 'cb',
n.report = 10,
n.burn = 100,
n.chains = 1)
expect_s3_class(out, "spMsPGOcc")
})
# Check default values ----------------
test_that("default priors, inits, burn, thin work", {
out <- spMsPGOcc(occ.formula = occ.formula,
det.formula = det.formula,
data = data.list,
n.batch = n.batch,
batch.length = batch.length,
accept.rate = 0.43,
cov.model = "exponential",
tuning = tuning.list,
NNGP = TRUE,
verbose = FALSE,
n.neighbors = 5,
n.chains = 1)
expect_s3_class(out, "spMsPGOcc")
})
test_that("all correlation functions work", {
out <- spMsPGOcc(occ.formula = occ.formula,
det.formula = det.formula,
data = data.list,
inits = inits.list,
n.batch = n.batch,
batch.length = batch.length,
accept.rate = 0.43,
priors = prior.list,
cov.model = "gaussian",
tuning = list(phi = 0.3),
n.omp.threads = 1,
verbose = FALSE,
NNGP = TRUE,
n.neighbors = 5,
search.type = 'cb',
n.report = 10,
n.burn = 100,
n.thin = 1,
n.chains = 1)
expect_s3_class(out, "spMsPGOcc")
out <- spMsPGOcc(occ.formula = occ.formula,
det.formula = det.formula,
data = data.list,
inits = inits.list,
n.batch = n.batch,
batch.length = batch.length,
accept.rate = 0.43,
priors = prior.list,
cov.model = "spherical",
tuning = list(phi = 0.3),
n.omp.threads = 1,
verbose = FALSE,
NNGP = TRUE,
n.neighbors = 5,
search.type = 'cb',
n.report = 10,
n.burn = 100,
n.thin = 1,
n.chains = 1)
expect_s3_class(out, "spMsPGOcc")
out <- spMsPGOcc(occ.formula = occ.formula,
det.formula = det.formula,
data = data.list,
inits = inits.list,
n.batch = n.batch,
batch.length = batch.length,
accept.rate = 0.43,
priors = list(nu.unif = list(0.4, 3)),
cov.model = "matern",
tuning = list(phi = 0.3, nu = 0.2),
n.omp.threads = 1,
verbose = FALSE,
NNGP = TRUE,
n.neighbors = 5,
search.type = 'cb',
n.report = 10,
n.burn = 100,
n.thin = 1,
n.chains = 1)
expect_s3_class(out, "spMsPGOcc")
})
test_that("verbose prints to the screen", {
expect_output(spMsPGOcc(occ.formula = occ.formula,
det.formula = det.formula,
data = data.list,
inits = inits.list,
n.batch = n.batch,
batch.length = batch.length,
accept.rate = 0.43,
priors = prior.list,
cov.model = "exponential",
tuning = tuning.list,
n.omp.threads = 1,
verbose = TRUE,
NNGP = TRUE,
n.neighbors = 5,
search.type = 'cb',
n.report = 10,
n.burn = 100,
n.thin = 1,
n.chains = 1))
})
# Check waicOcc -----------------------
test_that("waicOCC works for spMsPGOcc", {
# as.vector gets rid of names
waic.out <- as.vector(waicOcc(out))
expect_equal(length(waic.out), 3)
expect_equal(waic.out[3], -2 * (waic.out[1] - waic.out[2]))
})
test_that("fitted works for spMsPGOcc", {
fitted.out <- fitted(out)
expect_equal(length(fitted.out), 2)
expect_equal(class(fitted.out$y.rep.samples), "array")
expect_equal(class(fitted.out$p.samples), "array")
expect_equal(dim(fitted.out$y.rep.samples), dim(fitted.out$p.samples))
})
test_that("predict works for spMsPGOcc", {
n.post.samples <- out$n.post * out$n.chains
pred.out <- predict(out, X.0, coords.0, verbose = FALSE)
expect_type(pred.out, "list")
expect_equal(dim(pred.out$psi.0.samples), c(n.post.samples, N, nrow(X.0)))
expect_equal(dim(pred.out$z.0.samples), c(n.post.samples, N, nrow(X.0)))
})
test_that("detection prediction works", {
X.p.0 <- dat$X.p[, 1, ]
pred.out <- predict(out, X.p.0, type = 'detection')
expect_type(pred.out, 'list')
expect_equal(dim(pred.out$p.0.samples), c(out$n.post * out$n.chains, N, J))
})
test_that("posterior predictive checks work for msPGOcc", {
n.post.samples <- out$n.post * out$n.chains
J.fit <- nrow(X)
ppc.out <- ppcOcc(out, 'chi-square', 2)
expect_type(ppc.out, "list")
expect_equal(dim(ppc.out$fit.y), c(n.post.samples, N))
expect_equal(dim(ppc.out$fit.y.rep), c(n.post.samples, N))
expect_equal(dim(ppc.out$fit.y.group.quants), c(5, N, max(n.rep)))
expect_equal(dim(ppc.out$fit.y.rep.group.quants), c(5, N, max(n.rep)))
ppc.out <- ppcOcc(out, 'chi-square', 1)
expect_type(ppc.out, "list")
expect_equal(dim(ppc.out$fit.y), c(n.post.samples, N))
expect_equal(dim(ppc.out$fit.y.rep), c(n.post.samples, N))
expect_equal(dim(ppc.out$fit.y.group.quants), c(5, N, J.fit))
expect_equal(dim(ppc.out$fit.y.rep.group.quants), c(5, N, J.fit))
ppc.out <- ppcOcc(out, 'freeman-tukey', 1)
expect_type(ppc.out, "list")
expect_equal(dim(ppc.out$fit.y), c(n.post.samples, N))
expect_equal(dim(ppc.out$fit.y.rep), c(n.post.samples, N))
expect_equal(dim(ppc.out$fit.y.group.quants), c(5, N, J.fit))
expect_equal(dim(ppc.out$fit.y.rep.group.quants), c(5, N, J.fit))
ppc.out <- ppcOcc(out, 'freeman-tukey', 2)
expect_type(ppc.out, "list")
expect_equal(dim(ppc.out$fit.y), c(n.post.samples, N))
expect_equal(dim(ppc.out$fit.y.rep), c(n.post.samples, N))
expect_equal(dim(ppc.out$fit.y.group.quants), c(5, N, max(n.rep)))
expect_equal(dim(ppc.out$fit.y.rep.group.quants), c(5, N, max(n.rep)))
})
# Interactions on both ----------------------------------------------------
J.x <- 8
J.y <- 8
J <- J.x * J.y
n.rep<- sample(2:4, size = J, replace = TRUE)
N <- 6
# Community-level covariate effects
# Occurrence
beta.mean <- c(0.2, 0.5, 1.2)
p.occ <- length(beta.mean)
tau.sq.beta <- c(0.6, 1.5, 2.3)
# Detection
alpha.mean <- c(0, -0.5, 1.2)
tau.sq.alpha <- c(1, 2.3, 1.5)
p.det <- length(alpha.mean)
# Random effects
psi.RE <- list()
p.RE <- list()
# Draw species-level effects from community means.
beta <- matrix(NA, nrow = N, ncol = p.occ)
alpha <- matrix(NA, nrow = N, ncol = p.det)
for (i in 1:p.occ) {
beta[, i] <- rnorm(N, beta.mean[i], sqrt(tau.sq.beta[i]))
}
for (i in 1:p.det) {
alpha[, i] <- rnorm(N, alpha.mean[i], sqrt(tau.sq.alpha[i]))
}
alpha.true <- alpha
phi <- rep(3 / .7, N)
sigma.sq <- rep(2, N)
nu <- rep(2, N)
dat <- simMsOcc(J.x = J.x, J.y = J.y, n.rep = n.rep, N = N, beta = beta, alpha = alpha,
psi.RE = psi.RE, p.RE = p.RE, sp = TRUE, sigma.sq = sigma.sq,
phi = phi, nu = nu, cov.model = 'matern')
pred.indx <- sample(1:J, round(J * .25), replace = FALSE)
y <- dat$y[, -pred.indx, , drop = FALSE]
# Occupancy covariates
X <- dat$X[-pred.indx, , drop = FALSE]
coords <- as.matrix(dat$coords[-pred.indx, , drop = FALSE])
# Prediction covariates
X.0 <- dat$X[pred.indx, , drop = FALSE]
coords.0 <- as.matrix(dat$coords[pred.indx, , drop = FALSE])
# Detection covariates
X.p <- dat$X.p[-pred.indx, , , drop = FALSE]
occ.covs <- X
colnames(occ.covs) <- c('int', 'occ.cov.1', 'occ.cov.2')
det.covs <- list(det.cov.1 = X.p[, , 2],
det.cov.2 = X.p[, , 3])
data.list <- list(y = y, coords = coords, occ.covs = occ.covs, det.covs = det.covs)
# Priors
prior.list <- list(beta.comm.normal = list(mean = 0, var = 2.72),
alpha.comm.normal = list(mean = 0, var = 2.72),
tau.sq.beta.ig = list(a = 0.1, b = 0.1),
tau.sq.alpha.ig = list(a = 0.1, b = 0.1),
nu.unif = list(0.5, 2.5))
# Starting values
inits.list <- list(alpha.comm = 0,
beta.comm = 0,
beta = 0,
alpha = 0,
tau.sq.beta = 1,
tau.sq.alpha = 1,
z = apply(y, c(1, 2), max, na.rm = TRUE))
# Tuning
tuning.list <- list(phi = 1, nu = 0.25)
batch.length <- 25
n.batch <- 10
n.report <- 100
occ.formula <- ~ occ.cov.1 * occ.cov.2
det.formula <- ~ det.cov.1 * det.cov.2
out <- spMsPGOcc(occ.formula = occ.formula,
det.formula = det.formula,
data = data.list,
inits = inits.list,
n.batch = n.batch,
batch.length = batch.length,
accept.rate = 0.43,
priors = prior.list,
cov.model = "matern",
tuning = tuning.list,
n.omp.threads = 1,
verbose = FALSE,
NNGP = TRUE,
n.neighbors = 5,
search.type = 'cb',
n.report = 10,
n.burn = 100,
n.thin = 2,
n.chains = 2,
k.fold = 2,
k.fold.threads = 1)
# To make sure it worked --------------
test_that("out is of class spMsPGOcc", {
expect_s3_class(out, "spMsPGOcc")
})
# Check cross-validation --------------
test_that("cross-validation works", {
expect_equal(length(out$k.fold.deviance), N)
expect_type(out$k.fold.deviance, "double")
expect_equal(sum(out$k.fold.deviance < 0), 0)
})
# Check random effects ----------------
test_that("random effects are correct", {
expect_equal(out$pRE, FALSE)
expect_equal(out$psiRE, FALSE)
})
# Check output data output is correct -
test_that("out$y == y", {
expect_equal(out$y, y)
})
# Check missing values ----------------
test_that("missing value error handling works", {
tmp.data <- data.list
tmp.data$occ.covs[3, ] <- NA
expect_error(spMsPGOcc(occ.formula = occ.formula,
det.formula = det.formula,
data = tmp.data,
n.batch = 40,
batch.length = batch.length,
cov.model = "exponential",
tuning = tuning.list,
NNGP = TRUE,
verbose = FALSE,
n.neighbors = 5,
search.type = 'cb',
n.report = 10,
n.burn = 100,
n.chains = 1))
tmp.data <- data.list
tmp.data$y[1, 1, 1] <- 1
tmp.data$det.covs[[1]][1] <- NA
expect_error(spMsPGOcc(occ.formula = occ.formula,
det.formula = det.formula,
data = tmp.data,
n.batch = 40,
batch.length = batch.length,
cov.model = "exponential",
tuning = tuning.list,
NNGP = TRUE,
verbose = FALSE,
n.neighbors = 5,
search.type = 'cb',
n.report = 10,
n.burn = 100,
n.chains = 1))
tmp.data <- data.list
tmp.data$y[, 1, 1] <- NA
out <- spMsPGOcc(occ.formula = occ.formula,
det.formula = det.formula,
data = tmp.data,
n.batch = 40,
batch.length = batch.length,
cov.model = "exponential",
tuning = tuning.list,
NNGP = TRUE,
verbose = FALSE,
n.neighbors = 5,
search.type = 'cb',
n.report = 10,
n.burn = 100,
n.chains = 1)
expect_s3_class(out, "spMsPGOcc")
})
# Check default values ----------------
test_that("default priors, inits, burn, thin work", {
out <- spMsPGOcc(occ.formula = occ.formula,
det.formula = det.formula,
data = data.list,
n.batch = n.batch,
batch.length = batch.length,
accept.rate = 0.43,
cov.model = "exponential",
tuning = tuning.list,
NNGP = TRUE,
verbose = FALSE,
n.neighbors = 5,
n.chains = 1)
expect_s3_class(out, "spMsPGOcc")
})
test_that("all correlation functions work", {
out <- spMsPGOcc(occ.formula = occ.formula,
det.formula = det.formula,
data = data.list,
inits = inits.list,
n.batch = n.batch,
batch.length = batch.length,
accept.rate = 0.43,
priors = prior.list,
cov.model = "gaussian",
tuning = list(phi = 0.3),
n.omp.threads = 1,
verbose = FALSE,
NNGP = TRUE,
n.neighbors = 5,
search.type = 'cb',
n.report = 10,
n.burn = 100,
n.thin = 1,
n.chains = 1)
expect_s3_class(out, "spMsPGOcc")
out <- spMsPGOcc(occ.formula = occ.formula,
det.formula = det.formula,
data = data.list,
inits = inits.list,
n.batch = n.batch,
batch.length = batch.length,
accept.rate = 0.43,
priors = prior.list,
cov.model = "spherical",
tuning = list(phi = 0.3),
n.omp.threads = 1,
verbose = FALSE,
NNGP = TRUE,
n.neighbors = 5,
search.type = 'cb',
n.report = 10,
n.burn = 100,
n.thin = 1,
n.chains = 1)
expect_s3_class(out, "spMsPGOcc")
out <- spMsPGOcc(occ.formula = occ.formula,
det.formula = det.formula,
data = data.list,
inits = inits.list,
n.batch = n.batch,
batch.length = batch.length,
accept.rate = 0.43,
priors = list(nu.unif = list(0.4, 3)),
cov.model = "matern",
tuning = list(phi = 0.3, nu = 0.2),
n.omp.threads = 1,
verbose = FALSE,
NNGP = TRUE,
n.neighbors = 5,
search.type = 'cb',
n.report = 10,
n.burn = 100,
n.thin = 1,
n.chains = 1)
expect_s3_class(out, "spMsPGOcc")
})
test_that("verbose prints to the screen", {
expect_output(spMsPGOcc(occ.formula = occ.formula,
det.formula = det.formula,
data = data.list,
inits = inits.list,
n.batch = n.batch,
batch.length = batch.length,
accept.rate = 0.43,
priors = prior.list,
cov.model = "exponential",
tuning = tuning.list,
n.omp.threads = 1,
verbose = TRUE,
NNGP = TRUE,
n.neighbors = 5,
search.type = 'cb',
n.report = 10,
n.burn = 100,
n.thin = 1,
n.chains = 1))
})
# Check waicOcc -----------------------
test_that("waicOCC works for spMsPGOcc", {
# as.vector gets rid of names
waic.out <- as.vector(waicOcc(out))
expect_equal(length(waic.out), 3)
expect_equal(waic.out[3], -2 * (waic.out[1] - waic.out[2]))
})
test_that("fitted works for spMsPGOcc", {
fitted.out <- fitted(out)
expect_equal(length(fitted.out), 2)
expect_equal(class(fitted.out$y.rep.samples), "array")
expect_equal(class(fitted.out$p.samples), "array")
expect_equal(dim(fitted.out$y.rep.samples), dim(fitted.out$p.samples))
})
test_that("predict works for spMsPGOcc", {
n.post.samples <- out$n.post * out$n.chains
X.0 <- cbind(X.0, X.0[, 2] * X.0[, 3])
colnames(X.0) <- c('int', 'occ.cov.1', 'occ.cov.2', 'occ.cov.1:occ.cov.2')
pred.out <- predict(out, X.0, coords.0, verbose = FALSE)
expect_type(pred.out, "list")
expect_equal(dim(pred.out$psi.0.samples), c(n.post.samples, N, nrow(X.0)))
expect_equal(dim(pred.out$z.0.samples), c(n.post.samples, N, nrow(X.0)))
})
test_that("detection prediction works", {
X.p.0 <- dat$X.p[, 1, ]
X.p.0 <- cbind(X.p.0, X.p.0[, 2] * X.p.0[, 3])
pred.out <- predict(out, X.p.0, type = 'detection')
expect_type(pred.out, 'list')
expect_equal(dim(pred.out$p.0.samples), c(out$n.post * out$n.chains, N, J))
})
test_that("posterior predictive checks work for msPGOcc", {
n.post.samples <- out$n.post * out$n.chains
J.fit <- nrow(X)
ppc.out <- ppcOcc(out, 'chi-square', 2)
expect_type(ppc.out, "list")
expect_equal(dim(ppc.out$fit.y), c(n.post.samples, N))
expect_equal(dim(ppc.out$fit.y.rep), c(n.post.samples, N))
expect_equal(dim(ppc.out$fit.y.group.quants), c(5, N, max(n.rep)))
expect_equal(dim(ppc.out$fit.y.rep.group.quants), c(5, N, max(n.rep)))
ppc.out <- ppcOcc(out, 'chi-square', 1)
expect_type(ppc.out, "list")
expect_equal(dim(ppc.out$fit.y), c(n.post.samples, N))
expect_equal(dim(ppc.out$fit.y.rep), c(n.post.samples, N))
expect_equal(dim(ppc.out$fit.y.group.quants), c(5, N, J.fit))
expect_equal(dim(ppc.out$fit.y.rep.group.quants), c(5, N, J.fit))
ppc.out <- ppcOcc(out, 'freeman-tukey', 1)
expect_type(ppc.out, "list")
expect_equal(dim(ppc.out$fit.y), c(n.post.samples, N))
expect_equal(dim(ppc.out$fit.y.rep), c(n.post.samples, N))
expect_equal(dim(ppc.out$fit.y.group.quants), c(5, N, J.fit))
expect_equal(dim(ppc.out$fit.y.rep.group.quants), c(5, N, J.fit))
ppc.out <- ppcOcc(out, 'freeman-tukey', 2)
expect_type(ppc.out, "list")
expect_equal(dim(ppc.out$fit.y), c(n.post.samples, N))
expect_equal(dim(ppc.out$fit.y.rep), c(n.post.samples, N))
expect_equal(dim(ppc.out$fit.y.group.quants), c(5, N, max(n.rep)))
expect_equal(dim(ppc.out$fit.y.rep.group.quants), c(5, N, max(n.rep)))
})
# Site covariate on detection ---------------------------------------------
set.seed(400)
J.x <- 8
J.y <- 8
J <- J.x * J.y
n.rep<- sample(2:4, size = J, replace = TRUE)
N <- 6
# Community-level covariate effects
# Occurrence
beta.mean <- c(0.2, 0.5, 1.2)
p.occ <- length(beta.mean)
tau.sq.beta <- c(0.6, 1.5, 2.3)
# Detection
alpha.mean <- c(0, -0.5)
tau.sq.alpha <- c(1, 2.3)
p.det <- length(alpha.mean)
# Random effects
psi.RE <- list()
p.RE <- list()
# Draw species-level effects from community means.
beta <- matrix(NA, nrow = N, ncol = p.occ)
alpha <- matrix(NA, nrow = N, ncol = p.det)
for (i in 1:p.occ) {
beta[, i] <- rnorm(N, beta.mean[i], sqrt(tau.sq.beta[i]))
}
for (i in 1:p.det) {
alpha[, i] <- rnorm(N, alpha.mean[i], sqrt(tau.sq.alpha[i]))
}
alpha.true <- alpha
phi <- rep(3 / .7, N)
sigma.sq <- rep(2, N)
nu <- rep(2, N)
dat <- simMsOcc(J.x = J.x, J.y = J.y, n.rep = n.rep, N = N, beta = beta, alpha = alpha,
psi.RE = psi.RE, p.RE = p.RE, sp = TRUE, sigma.sq = sigma.sq,
phi = phi, nu = nu, cov.model = 'matern')
pred.indx <- sample(1:J, round(J * .25), replace = FALSE)
y <- dat$y[, -pred.indx, , drop = FALSE]
# Occupancy covariates
X <- dat$X[-pred.indx, , drop = FALSE]
coords <- as.matrix(dat$coords[-pred.indx, , drop = FALSE])
# Prediction covariates
X.0 <- dat$X[pred.indx, , drop = FALSE]
coords.0 <- as.matrix(dat$coords[pred.indx, , drop = FALSE])
# Detection covariates
X.p <- dat$X.p[-pred.indx, , , drop = FALSE]
occ.covs <- X
colnames(occ.covs) <- c('int', 'occ.cov.1', 'occ.cov.2')
det.covs <- list(det.cov.1 = X.p[, , 2],
occ.cov.1 = X[, 2])
data.list <- list(y = y, coords = coords, occ.covs = occ.covs, det.covs = det.covs)
# Priors
prior.list <- list(beta.comm.normal = list(mean = 0, var = 2.72),
alpha.comm.normal = list(mean = 0, var = 2.72),
tau.sq.beta.ig = list(a = 0.1, b = 0.1),
tau.sq.alpha.ig = list(a = 0.1, b = 0.1),
nu.unif = list(0.5, 2.5))
# Starting values
inits.list <- list(alpha.comm = 0,
beta.comm = 0,
beta = 0,
alpha = 0,
tau.sq.beta = 1,
tau.sq.alpha = 1,
z = apply(y, c(1, 2), max, na.rm = TRUE))
# Tuning
tuning.list <- list(phi = 1, nu = 0.25)
batch.length <- 25
n.batch <- 10
n.report <- 100
occ.formula <- ~ 1
det.formula <- ~ occ.cov.1
out <- spMsPGOcc(occ.formula = occ.formula,
det.formula = det.formula,
data = data.list,
inits = inits.list,
n.batch = n.batch,
batch.length = batch.length,
accept.rate = 0.43,
priors = prior.list,
cov.model = "matern",
tuning = tuning.list,
n.omp.threads = 1,
verbose = FALSE,
NNGP = TRUE,
n.neighbors = 5,
search.type = 'cb',
n.report = 10,
n.burn = 100,
n.thin = 2,
n.chains = 2,
k.fold = 2,
k.fold.threads = 1)
# To make sure it worked --------------
test_that("out is of class spMsPGOcc", {
expect_s3_class(out, "spMsPGOcc")
})
# Check cross-validation --------------
test_that("cross-validation works", {
expect_equal(length(out$k.fold.deviance), N)
expect_type(out$k.fold.deviance, "double")
expect_equal(sum(out$k.fold.deviance < 0), 0)
})
# Check random effects ----------------
test_that("random effects are correct", {
expect_equal(out$pRE, FALSE)
expect_equal(out$psiRE, FALSE)
})
# Check output data output is correct -
test_that("out$y == y", {
expect_equal(out$y, y)
})
# Check missing values ----------------
test_that("missing value error handling works", {
# tmp.data <- data.list
# tmp.data$occ.covs[3, ] <- NA
# expect_error(spMsPGOcc(occ.formula = occ.formula,
# det.formula = det.formula,
# data = tmp.data,
# n.batch = 40,
# batch.length = batch.length,
# cov.model = "exponential",
# tuning = tuning.list,
# NNGP = TRUE,
# verbose = FALSE,
# n.neighbors = 5,
# search.type = 'cb',
# n.report = 10,
# n.burn = 100,
# n.chains = 1))
tmp.data <- data.list
tmp.data$y[1, 1, 1] <- 1
tmp.data$det.covs[[2]][1] <- NA
expect_error(spMsPGOcc(occ.formula = occ.formula,
det.formula = det.formula,
data = tmp.data,
n.batch = 40,
batch.length = batch.length,
cov.model = "exponential",
tuning = tuning.list,
NNGP = TRUE,
verbose = FALSE,
n.neighbors = 5,
search.type = 'cb',
n.report = 10,
n.burn = 100,
n.chains = 1))
tmp.data <- data.list
tmp.data$y[, 1, 1] <- NA
out <- spMsPGOcc(occ.formula = occ.formula,
det.formula = det.formula,
data = tmp.data,
n.batch = 40,
batch.length = batch.length,
cov.model = "exponential",
tuning = tuning.list,
NNGP = TRUE,
verbose = FALSE,
n.neighbors = 5,
search.type = 'cb',
n.report = 10,
n.burn = 100,
n.chains = 1)
expect_s3_class(out, "spMsPGOcc")
})
# Check default values ----------------
test_that("default priors, inits, burn, thin work", {
out <- spMsPGOcc(occ.formula = occ.formula,
det.formula = det.formula,
data = data.list,
n.batch = n.batch,
batch.length = batch.length,
accept.rate = 0.43,
cov.model = "exponential",
tuning = tuning.list,
NNGP = TRUE,
verbose = FALSE,
n.neighbors = 5,
n.chains = 1)
expect_s3_class(out, "spMsPGOcc")
})
test_that("all correlation functions work", {
out <- spMsPGOcc(occ.formula = occ.formula,
det.formula = det.formula,
data = data.list,
inits = inits.list,
n.batch = n.batch,
batch.length = batch.length,
accept.rate = 0.43,
priors = prior.list,
cov.model = "gaussian",
tuning = list(phi = 0.3),
n.omp.threads = 1,
verbose = FALSE,
NNGP = TRUE,
n.neighbors = 5,
search.type = 'cb',
n.report = 10,
n.burn = 100,
n.thin = 1,
n.chains = 1)
expect_s3_class(out, "spMsPGOcc")
out <- spMsPGOcc(occ.formula = occ.formula,
det.formula = det.formula,
data = data.list,
inits = inits.list,
n.batch = n.batch,
batch.length = batch.length,
accept.rate = 0.43,
priors = prior.list,
cov.model = "spherical",
tuning = list(phi = 0.3),
n.omp.threads = 1,
verbose = FALSE,
NNGP = TRUE,
n.neighbors = 5,
search.type = 'cb',
n.report = 10,
n.burn = 100,
n.thin = 1,
n.chains = 1)
expect_s3_class(out, "spMsPGOcc")
out <- spMsPGOcc(occ.formula = occ.formula,
det.formula = det.formula,
data = data.list,
inits = inits.list,
n.batch = n.batch,
batch.length = batch.length,
accept.rate = 0.43,
priors = list(nu.unif = list(0.4, 3)),
cov.model = "matern",
tuning = list(phi = 0.3, nu = 0.2),
n.omp.threads = 1,
verbose = FALSE,
NNGP = TRUE,
n.neighbors = 5,
search.type = 'cb',
n.report = 10,
n.burn = 100,
n.thin = 1,
n.chains = 1)
expect_s3_class(out, "spMsPGOcc")
})
test_that("verbose prints to the screen", {
expect_output(spMsPGOcc(occ.formula = occ.formula,
det.formula = det.formula,
data = data.list,
inits = inits.list,
n.batch = n.batch,
batch.length = batch.length,
accept.rate = 0.43,
priors = prior.list,
cov.model = "exponential",
tuning = tuning.list,
n.omp.threads = 1,
verbose = TRUE,
NNGP = TRUE,
n.neighbors = 5,
search.type = 'cb',
n.report = 10,
n.burn = 100,
n.thin = 1,
n.chains = 1))
})
# Check waicOcc -----------------------
test_that("waicOCC works for spMsPGOcc", {
# as.vector gets rid of names
waic.out <- as.vector(waicOcc(out))
expect_equal(length(waic.out), 3)
expect_equal(waic.out[3], -2 * (waic.out[1] - waic.out[2]))
})
test_that("fitted works for spMsPGOcc", {
fitted.out <- fitted(out)
expect_equal(length(fitted.out), 2)
expect_equal(class(fitted.out$y.rep.samples), "array")
expect_equal(class(fitted.out$p.samples), "array")
expect_equal(dim(fitted.out$y.rep.samples), dim(fitted.out$p.samples))
})
test_that("predict works for spMsPGOcc", {
n.post.samples <- out$n.post * out$n.chains
X.0 <- X.0[, 1, drop = FALSE]
colnames(X.0) <- c('int')
pred.out <- predict(out, X.0, coords.0, verbose = FALSE)
expect_type(pred.out, "list")
expect_equal(dim(pred.out$psi.0.samples), c(n.post.samples, N, nrow(X.0)))
expect_equal(dim(pred.out$z.0.samples), c(n.post.samples, N, nrow(X.0)))
})
test_that("detection prediction works", {
X.p.0 <- cbind(1, data.list$det.covs$occ.cov.1)
pred.out <- predict(out, X.p.0, type = 'detection')
expect_type(pred.out, 'list')
expect_equal(dim(pred.out$p.0.samples), c(out$n.post * out$n.chains, N, nrow(X.p.0)))
})
test_that("posterior predictive checks work for msPGOcc", {
n.post.samples <- out$n.post * out$n.chains
J.fit <- nrow(X)
ppc.out <- ppcOcc(out, 'chi-square', 2)
expect_type(ppc.out, "list")
expect_equal(dim(ppc.out$fit.y), c(n.post.samples, N))
expect_equal(dim(ppc.out$fit.y.rep), c(n.post.samples, N))
expect_equal(dim(ppc.out$fit.y.group.quants), c(5, N, max(n.rep)))
expect_equal(dim(ppc.out$fit.y.rep.group.quants), c(5, N, max(n.rep)))
ppc.out <- ppcOcc(out, 'chi-square', 1)
expect_type(ppc.out, "list")
expect_equal(dim(ppc.out$fit.y), c(n.post.samples, N))
expect_equal(dim(ppc.out$fit.y.rep), c(n.post.samples, N))
expect_equal(dim(ppc.out$fit.y.group.quants), c(5, N, J.fit))
expect_equal(dim(ppc.out$fit.y.rep.group.quants), c(5, N, J.fit))
ppc.out <- ppcOcc(out, 'freeman-tukey', 1)
expect_type(ppc.out, "list")
expect_equal(dim(ppc.out$fit.y), c(n.post.samples, N))
expect_equal(dim(ppc.out$fit.y.rep), c(n.post.samples, N))
expect_equal(dim(ppc.out$fit.y.group.quants), c(5, N, J.fit))
expect_equal(dim(ppc.out$fit.y.rep.group.quants), c(5, N, J.fit))
ppc.out <- ppcOcc(out, 'freeman-tukey', 2)
expect_type(ppc.out, "list")
expect_equal(dim(ppc.out$fit.y), c(n.post.samples, N))
expect_equal(dim(ppc.out$fit.y.rep), c(n.post.samples, N))
expect_equal(dim(ppc.out$fit.y.group.quants), c(5, N, max(n.rep)))
expect_equal(dim(ppc.out$fit.y.rep.group.quants), c(5, N, max(n.rep)))
})
# Random intercept on occurrence -------------------------------------------
J.x <- 8
J.y <- 8
J <- J.x * J.y
n.rep<- sample(2:4, size = J, replace = TRUE)
N <- 6
# Community-level covariate effects
# Occurrence
beta.mean <- c(0.2)
p.occ <- length(beta.mean)
tau.sq.beta <- c(0.6)
# Detection
alpha.mean <- c(0)
tau.sq.alpha <- c(1)
p.det <- length(alpha.mean)
# Random effects
psi.RE <- list(levels = c(45),
sigma.sq.psi = c(1.3))
p.RE <- list()
# Draw species-level effects from community means.
beta <- matrix(NA, nrow = N, ncol = p.occ)
alpha <- matrix(NA, nrow = N, ncol = p.det)
for (i in 1:p.occ) {
beta[, i] <- rnorm(N, beta.mean[i], sqrt(tau.sq.beta[i]))
}
for (i in 1:p.det) {
alpha[, i] <- rnorm(N, alpha.mean[i], sqrt(tau.sq.alpha[i]))
}
alpha.true <- alpha
phi <- rep(3 / .7, N)
sigma.sq <- rep(2, N)
nu <- rep(2, N)
dat <- simMsOcc(J.x = J.x, J.y = J.y, n.rep = n.rep, N = N, beta = beta, alpha = alpha,
psi.RE = psi.RE, p.RE = p.RE, sp = TRUE, sigma.sq = sigma.sq,
phi = phi, nu = nu, cov.model = 'matern')
pred.indx <- sample(1:J, round(J * .25), replace = FALSE)
y <- dat$y[, -pred.indx, , drop = FALSE]
# Occupancy covariates
X <- dat$X[-pred.indx, , drop = FALSE]
X.re <- dat$X.re[-pred.indx, , drop = FALSE]
coords <- as.matrix(dat$coords[-pred.indx, , drop = FALSE])
# Prediction covariates
X.0 <- dat$X[pred.indx, , drop = FALSE]
X.re.0 <- dat$X.re[pred.indx, , drop = FALSE]
coords.0 <- as.matrix(dat$coords[pred.indx, , drop = FALSE])
# Detection covariates
X.p <- dat$X.p[-pred.indx, , , drop = FALSE]
occ.covs <- cbind(X, X.re)
colnames(occ.covs) <- c('int', 'occ.factor.1')
data.list <- list(y = y, coords = coords, occ.covs = occ.covs)
# Priors
prior.list <- list(beta.comm.normal = list(mean = 0, var = 2.72),
alpha.comm.normal = list(mean = 0, var = 2.72),
tau.sq.beta.ig = list(a = 0.1, b = 0.1),
tau.sq.alpha.ig = list(a = 0.1, b = 0.1),
nu.unif = list(0.5, 2.5))
# Starting values
inits.list <- list(alpha.comm = 0,
beta.comm = 0,
beta = 0,
alpha = 0,
tau.sq.beta = 1,
tau.sq.alpha = 1,
z = apply(y, c(1, 2), max, na.rm = TRUE))
# Tuning
tuning.list <- list(phi = 1, nu = 0.25)
batch.length <- 25
n.batch <- 10
n.report <- 100
occ.formula <- ~ (1 | occ.factor.1)
det.formula <- ~ 1
out <- spMsPGOcc(occ.formula = occ.formula,
det.formula = det.formula,
data = data.list,
inits = inits.list,
n.batch = n.batch,
batch.length = batch.length,
accept.rate = 0.43,
priors = prior.list,
cov.model = "matern",
tuning = tuning.list,
n.omp.threads = 1,
verbose = FALSE,
NNGP = TRUE,
n.neighbors = 5,
search.type = 'cb',
n.report = 10,
n.burn = 100,
n.thin = 2,
n.chains = 2,
k.fold = 2,
k.fold.threads = 1)
# To make sure it worked --------------
test_that("out is of class spMsPGOcc", {
expect_s3_class(out, "spMsPGOcc")
})
# Check cross-validation --------------
test_that("cross-validation works", {
expect_equal(length(out$k.fold.deviance), N)
expect_type(out$k.fold.deviance, "double")
expect_equal(sum(out$k.fold.deviance < 0), 0)
})
# Check random effects ----------------
test_that("random effects are correct", {
expect_equal(out$pRE, FALSE)
expect_equal(out$psiRE, TRUE)
})
# Check RE error ----------------------
test_that("random effect gives error when non-numeric", {
data.list$occ.covs <- as.data.frame(data.list$occ.covs)
data.list$occ.covs$occ.factor.1 <- factor(data.list$occ.covs$occ.factor.1)
# data.list$det.covs$det.factor.1 <- factor(data.list$det.covs$det.factor.1)
expect_error(out <- spMsPGOcc(occ.formula = occ.formula,
det.formula = det.formula,
data = data.list,
inits = inits.list,
n.batch = n.batch,
batch.length = batch.length,
accept.rate = 0.43,
priors = prior.list,
cov.model = "matern",
tuning = tuning.list,
n.omp.threads = 1,
verbose = FALSE,
NNGP = TRUE,
n.neighbors = 5,
search.type = 'cb',
n.report = 10,
n.burn = 100,
n.thin = 2,
n.chains = 1))
data.list$occ.covs$occ.factor.1 <- as.character(factor(data.list$occ.covs$occ.factor.1))
# data.list$det.covs$det.factor.1 <- as.character(factor(data.list$det.covs$det.factor.1))
expect_error(out <- spMsPGOcc(occ.formula = occ.formula,
det.formula = det.formula,
data = data.list,
inits = inits.list,
n.batch = n.batch,
batch.length = batch.length,
accept.rate = 0.43,
priors = prior.list,
cov.model = "matern",
tuning = tuning.list,
n.omp.threads = 1,
verbose = FALSE,
NNGP = TRUE,
n.neighbors = 5,
search.type = 'cb',
n.report = 10,
n.burn = 100,
n.thin = 2,
n.chains = 1))
})
# Check RE levels ---------------------
test_that("random effect levels are correct", {
expect_equal(sort(unique(unlist(out$re.level.names))),
sort(unique(c(X.re))))
})
# Check output data output is correct -
test_that("out$y == y", {
expect_equal(out$y, y)
})
# Check missing values ----------------
test_that("missing value error handling works", {
tmp.data <- data.list
tmp.data$occ.covs[3, ] <- NA
expect_error(spMsPGOcc(occ.formula = occ.formula,
det.formula = det.formula,
data = tmp.data,
n.batch = 40,
batch.length = batch.length,
cov.model = "exponential",
tuning = tuning.list,
NNGP = TRUE,
verbose = FALSE,
n.neighbors = 5,
search.type = 'cb',
n.report = 10,
n.burn = 100,
n.chains = 1))
# tmp.data <- data.list
# tmp.data$det.covs[[1]][1] <- NA
# expect_error(spMsPGOcc(occ.formula = occ.formula,
# det.formula = det.formula,
# data = tmp.data,
# n.batch = 40,
# batch.length = batch.length,
# cov.model = "exponential",
# tuning = tuning.list,
# NNGP = TRUE,
# verbose = FALSE,
# n.neighbors = 5,
# search.type = 'cb',
# n.report = 10,
# n.burn = 100,
# n.chains = 1))
# tmp.data <- data.list
# tmp.data$y[, 1, 1] <- NA
# out <- spMsPGOcc(occ.formula = occ.formula,
# det.formula = det.formula,
# data = tmp.data,
# n.batch = 40,
# batch.length = batch.length,
# cov.model = "exponential",
# tuning = tuning.list,
# NNGP = TRUE,
# verbose = FALSE,
# n.neighbors = 5,
# search.type = 'cb',
# n.report = 10,
# n.burn = 100,
# n.chains = 1)
# expect_s3_class(out, "spMsPGOcc")
})
# Check default values ----------------
test_that("default priors, inits, burn, thin work", {
out <- spMsPGOcc(occ.formula = occ.formula,
det.formula = det.formula,
data = data.list,
n.batch = n.batch,
batch.length = batch.length,
accept.rate = 0.43,
cov.model = "exponential",
tuning = tuning.list,
NNGP = TRUE,
verbose = FALSE,
n.neighbors = 5,
n.chains = 1)
expect_s3_class(out, "spMsPGOcc")
})
test_that("all correlation functions work", {
out <- spMsPGOcc(occ.formula = occ.formula,
det.formula = det.formula,
data = data.list,
inits = inits.list,
n.batch = n.batch,
batch.length = batch.length,
accept.rate = 0.43,
priors = prior.list,
cov.model = "gaussian",
tuning = list(phi = 0.3),
n.omp.threads = 1,
verbose = FALSE,
NNGP = TRUE,
n.neighbors = 5,
search.type = 'cb',
n.report = 10,
n.burn = 100,
n.thin = 1,
n.chains = 1)
expect_s3_class(out, "spMsPGOcc")
out <- spMsPGOcc(occ.formula = occ.formula,
det.formula = det.formula,
data = data.list,
inits = inits.list,
n.batch = n.batch,
batch.length = batch.length,
accept.rate = 0.43,
priors = prior.list,
cov.model = "spherical",
tuning = list(phi = 0.3),
n.omp.threads = 1,
verbose = FALSE,
NNGP = TRUE,
n.neighbors = 5,
search.type = 'cb',
n.report = 10,
n.burn = 100,
n.thin = 1,
n.chains = 1)
expect_s3_class(out, "spMsPGOcc")
out <- spMsPGOcc(occ.formula = occ.formula,
det.formula = det.formula,
data = data.list,
inits = inits.list,
n.batch = n.batch,
batch.length = batch.length,
accept.rate = 0.43,
priors = list(nu.unif = list(0.4, 3)),
cov.model = "matern",
tuning = list(phi = 0.3, nu = 0.2),
n.omp.threads = 1,
verbose = FALSE,
NNGP = TRUE,
n.neighbors = 5,
search.type = 'cb',
n.report = 10,
n.burn = 100,
n.thin = 1,
n.chains = 1)
expect_s3_class(out, "spMsPGOcc")
})
test_that("verbose prints to the screen", {
expect_output(spMsPGOcc(occ.formula = occ.formula,
det.formula = det.formula,
data = data.list,
inits = inits.list,
n.batch = n.batch,
batch.length = batch.length,
accept.rate = 0.43,
priors = prior.list,
cov.model = "exponential",
tuning = tuning.list,
n.omp.threads = 1,
verbose = TRUE,
NNGP = TRUE,
n.neighbors = 5,
search.type = 'cb',
n.report = 10,
n.burn = 100,
n.thin = 1,
n.chains = 1))
})
# Check waicOcc -----------------------
test_that("waicOCC works for spMsPGOcc", {
# as.vector gets rid of names
waic.out <- as.vector(waicOcc(out))
expect_equal(length(waic.out), 3)
expect_equal(waic.out[3], -2 * (waic.out[1] - waic.out[2]))
})
test_that("fitted works for spMsPGOcc", {
fitted.out <- fitted(out)
expect_equal(length(fitted.out), 2)
expect_equal(class(fitted.out$y.rep.samples), "array")
expect_equal(class(fitted.out$p.samples), "array")
expect_equal(dim(fitted.out$y.rep.samples), dim(fitted.out$p.samples))
})
test_that("predict works for spMsPGOcc", {
n.post.samples <- out$n.post * out$n.chains
X.0 <- cbind(X.0, X.re.0)
colnames(X.0) <- c('int', 'occ.factor.1')
pred.out <- predict(out, X.0, coords.0, verbose = FALSE)
expect_type(pred.out, "list")
expect_equal(dim(pred.out$psi.0.samples), c(n.post.samples, N, nrow(X.0)))
expect_equal(dim(pred.out$z.0.samples), c(n.post.samples, N, nrow(X.0)))
})
test_that("detection prediction works", {
J.str <- 100
X.p.0 <- matrix(1, nrow = J.str, ncol = p.det)
pred.out <- predict(out, X.p.0, type = 'detection')
expect_type(pred.out, 'list')
expect_equal(dim(pred.out$p.0.samples), c(out$n.post * out$n.chains, N, J.str))
})
test_that("posterior predictive checks work for msPGOcc", {
n.post.samples <- out$n.post * out$n.chains
J.fit <- nrow(X)
ppc.out <- ppcOcc(out, 'chi-square', 2)
expect_type(ppc.out, "list")
expect_equal(dim(ppc.out$fit.y), c(n.post.samples, N))
expect_equal(dim(ppc.out$fit.y.rep), c(n.post.samples, N))
expect_equal(dim(ppc.out$fit.y.group.quants), c(5, N, max(n.rep)))
expect_equal(dim(ppc.out$fit.y.rep.group.quants), c(5, N, max(n.rep)))
ppc.out <- ppcOcc(out, 'chi-square', 1)
expect_type(ppc.out, "list")
expect_equal(dim(ppc.out$fit.y), c(n.post.samples, N))
expect_equal(dim(ppc.out$fit.y.rep), c(n.post.samples, N))
expect_equal(dim(ppc.out$fit.y.group.quants), c(5, N, J.fit))
expect_equal(dim(ppc.out$fit.y.rep.group.quants), c(5, N, J.fit))
ppc.out <- ppcOcc(out, 'freeman-tukey', 1)
expect_type(ppc.out, "list")
expect_equal(dim(ppc.out$fit.y), c(n.post.samples, N))
expect_equal(dim(ppc.out$fit.y.rep), c(n.post.samples, N))
expect_equal(dim(ppc.out$fit.y.group.quants), c(5, N, J.fit))
expect_equal(dim(ppc.out$fit.y.rep.group.quants), c(5, N, J.fit))
ppc.out <- ppcOcc(out, 'freeman-tukey', 2)
expect_type(ppc.out, "list")
expect_equal(dim(ppc.out$fit.y), c(n.post.samples, N))
expect_equal(dim(ppc.out$fit.y.rep), c(n.post.samples, N))
expect_equal(dim(ppc.out$fit.y.group.quants), c(5, N, max(n.rep)))
expect_equal(dim(ppc.out$fit.y.rep.group.quants), c(5, N, max(n.rep)))
})
# Multiple random intercepts on occurrence --------------------------------
J.x <- 8
J.y <- 8
J <- J.x * J.y
n.rep<- sample(2:4, size = J, replace = TRUE)
N <- 6
# Community-level covariate effects
# Occurrence
beta.mean <- c(0.2)
p.occ <- length(beta.mean)
tau.sq.beta <- c(0.6)
# Detection
alpha.mean <- c(0)
tau.sq.alpha <- c(1)
p.det <- length(alpha.mean)
# Random effects
psi.RE <- list(levels = c(45, 20),
sigma.sq.psi = c(1.3, 3.4))
p.RE <- list()
# Draw species-level effects from community means.
beta <- matrix(NA, nrow = N, ncol = p.occ)
alpha <- matrix(NA, nrow = N, ncol = p.det)
for (i in 1:p.occ) {
beta[, i] <- rnorm(N, beta.mean[i], sqrt(tau.sq.beta[i]))
}
for (i in 1:p.det) {
alpha[, i] <- rnorm(N, alpha.mean[i], sqrt(tau.sq.alpha[i]))
}
alpha.true <- alpha
phi <- rep(3 / .7, N)
sigma.sq <- rep(2, N)
nu <- rep(2, N)
dat <- simMsOcc(J.x = J.x, J.y = J.y, n.rep = n.rep, N = N, beta = beta, alpha = alpha,
psi.RE = psi.RE, p.RE = p.RE, sp = TRUE, sigma.sq = sigma.sq,
phi = phi, nu = nu, cov.model = 'matern')
pred.indx <- sample(1:J, round(J * .25), replace = FALSE)
y <- dat$y[, -pred.indx, , drop = FALSE]
# Occupancy covariates
X <- dat$X[-pred.indx, , drop = FALSE]
X.re <- dat$X.re[-pred.indx, , drop = FALSE]
coords <- as.matrix(dat$coords[-pred.indx, , drop = FALSE])
# Prediction covariates
X.0 <- dat$X[pred.indx, , drop = FALSE]
X.re.0 <- dat$X.re[pred.indx, , drop = FALSE]
coords.0 <- as.matrix(dat$coords[pred.indx, , drop = FALSE])
# Detection covariates
X.p <- dat$X.p[-pred.indx, , , drop = FALSE]
occ.covs <- cbind(X, X.re)
colnames(occ.covs) <- c('int', 'occ.factor.1', 'occ.factor.2')
data.list <- list(y = y, coords = coords, occ.covs = occ.covs)
# Priors
prior.list <- list(beta.comm.normal = list(mean = 0, var = 2.72),
alpha.comm.normal = list(mean = 0, var = 2.72),
tau.sq.beta.ig = list(a = 0.1, b = 0.1),
tau.sq.alpha.ig = list(a = 0.1, b = 0.1),
nu.unif = list(0.5, 2.5))
# Starting values
inits.list <- list(alpha.comm = 0,
beta.comm = 0,
beta = 0,
alpha = 0,
tau.sq.beta = 1,
tau.sq.alpha = 1,
z = apply(y, c(1, 2), max, na.rm = TRUE))
# Tuning
tuning.list <- list(phi = 1, nu = 0.25)
batch.length <- 25
n.batch <- 10
n.report <- 100
occ.formula <- ~ (1 | occ.factor.1) + (1 | occ.factor.2)
det.formula <- ~ 1
out <- spMsPGOcc(occ.formula = occ.formula,
det.formula = det.formula,
data = data.list,
inits = inits.list,
n.batch = n.batch,
batch.length = batch.length,
accept.rate = 0.43,
priors = prior.list,
cov.model = "matern",
tuning = tuning.list,
n.omp.threads = 1,
verbose = FALSE,
NNGP = TRUE,
n.neighbors = 5,
search.type = 'cb',
n.report = 10,
n.burn = 100,
n.thin = 2,
n.chains = 2,
k.fold = 2,
k.fold.threads = 1)
# To make sure it worked --------------
test_that("out is of class spMsPGOcc", {
expect_s3_class(out, "spMsPGOcc")
})
# Check cross-validation --------------
test_that("cross-validation works", {
expect_equal(length(out$k.fold.deviance), N)
expect_type(out$k.fold.deviance, "double")
expect_equal(sum(out$k.fold.deviance < 0), 0)
})
# Check random effects ----------------
test_that("random effects are correct", {
expect_equal(out$pRE, FALSE)
expect_equal(out$psiRE, TRUE)
})
# Check RE error ----------------------
test_that("random effect gives error when non-numeric", {
data.list$occ.covs <- as.data.frame(data.list$occ.covs)
data.list$occ.covs$occ.factor.1 <- factor(data.list$occ.covs$occ.factor.1)
# data.list$det.covs$det.factor.1 <- factor(data.list$det.covs$det.factor.1)
expect_error(out <- spMsPGOcc(occ.formula = occ.formula,
det.formula = det.formula,
data = data.list,
inits = inits.list,
n.batch = n.batch,
batch.length = batch.length,
accept.rate = 0.43,
priors = prior.list,
cov.model = "matern",
tuning = tuning.list,
n.omp.threads = 1,
verbose = FALSE,
NNGP = TRUE,
n.neighbors = 5,
search.type = 'cb',
n.report = 10,
n.burn = 100,
n.thin = 2,
n.chains = 1))
data.list$occ.covs$occ.factor.1 <- as.character(factor(data.list$occ.covs$occ.factor.1))
# data.list$det.covs$det.factor.1 <- as.character(factor(data.list$det.covs$det.factor.1))
expect_error(out <- spMsPGOcc(occ.formula = occ.formula,
det.formula = det.formula,
data = data.list,
inits = inits.list,
n.batch = n.batch,
batch.length = batch.length,
accept.rate = 0.43,
priors = prior.list,
cov.model = "matern",
tuning = tuning.list,
n.omp.threads = 1,
verbose = FALSE,
NNGP = TRUE,
n.neighbors = 5,
search.type = 'cb',
n.report = 10,
n.burn = 100,
n.thin = 2,
n.chains = 1))
})
# Check RE levels ---------------------
test_that("random effect levels are correct", {
expect_equal(sort(unique(unlist(out$re.level.names))),
sort(unique(c(X.re))))
})
# Check output data output is correct -
test_that("out$y == y", {
expect_equal(out$y, y)
})
# Check missing values ----------------
test_that("missing value error handling works", {
tmp.data <- data.list
tmp.data$occ.covs[3, ] <- NA
expect_error(spMsPGOcc(occ.formula = occ.formula,
det.formula = det.formula,
data = tmp.data,
n.batch = 40,
batch.length = batch.length,
cov.model = "exponential",
tuning = tuning.list,
NNGP = TRUE,
verbose = FALSE,
n.neighbors = 5,
search.type = 'cb',
n.report = 10,
n.burn = 100,
n.chains = 1))
# tmp.data <- data.list
# tmp.data$det.covs[[1]][1] <- NA
# expect_error(spMsPGOcc(occ.formula = occ.formula,
# det.formula = det.formula,
# data = tmp.data,
# n.batch = 40,
# batch.length = batch.length,
# cov.model = "exponential",
# tuning = tuning.list,
# NNGP = TRUE,
# verbose = FALSE,
# n.neighbors = 5,
# search.type = 'cb',
# n.report = 10,
# n.burn = 100,
# n.chains = 1))
# tmp.data <- data.list
# tmp.data$y[, 1, 1] <- NA
# out <- spMsPGOcc(occ.formula = occ.formula,
# det.formula = det.formula,
# data = tmp.data,
# n.batch = 40,
# batch.length = batch.length,
# cov.model = "exponential",
# tuning = tuning.list,
# NNGP = TRUE,
# verbose = FALSE,
# n.neighbors = 5,
# search.type = 'cb',
# n.report = 10,
# n.burn = 100,
# n.chains = 1)
# expect_s3_class(out, "spMsPGOcc")
})
# Check default values ----------------
test_that("default priors, inits, burn, thin work", {
out <- spMsPGOcc(occ.formula = occ.formula,
det.formula = det.formula,
data = data.list,
n.batch = n.batch,
batch.length = batch.length,
accept.rate = 0.43,
cov.model = "exponential",
tuning = tuning.list,
NNGP = TRUE,
verbose = FALSE,
n.neighbors = 5,
n.chains = 1)
expect_s3_class(out, "spMsPGOcc")
})
test_that("all correlation functions work", {
out <- spMsPGOcc(occ.formula = occ.formula,
det.formula = det.formula,
data = data.list,
inits = inits.list,
n.batch = n.batch,
batch.length = batch.length,
accept.rate = 0.43,
priors = prior.list,
cov.model = "gaussian",
tuning = list(phi = 0.3),
n.omp.threads = 1,
verbose = FALSE,
NNGP = TRUE,
n.neighbors = 5,
search.type = 'cb',
n.report = 10,
n.burn = 100,
n.thin = 1,
n.chains = 1)
expect_s3_class(out, "spMsPGOcc")
out <- spMsPGOcc(occ.formula = occ.formula,
det.formula = det.formula,
data = data.list,
inits = inits.list,
n.batch = n.batch,
batch.length = batch.length,
accept.rate = 0.43,
priors = prior.list,
cov.model = "spherical",
tuning = list(phi = 0.3),
n.omp.threads = 1,
verbose = FALSE,
NNGP = TRUE,
n.neighbors = 5,
search.type = 'cb',
n.report = 10,
n.burn = 100,
n.thin = 1,
n.chains = 1)
expect_s3_class(out, "spMsPGOcc")
out <- spMsPGOcc(occ.formula = occ.formula,
det.formula = det.formula,
data = data.list,
inits = inits.list,
n.batch = n.batch,
batch.length = batch.length,
accept.rate = 0.43,
priors = list(nu.unif = list(0.4, 3)),
cov.model = "matern",
tuning = list(phi = 0.3, nu = 0.2),
n.omp.threads = 1,
verbose = FALSE,
NNGP = TRUE,
n.neighbors = 5,
search.type = 'cb',
n.report = 10,
n.burn = 100,
n.thin = 1,
n.chains = 1)
expect_s3_class(out, "spMsPGOcc")
})
test_that("verbose prints to the screen", {
expect_output(spMsPGOcc(occ.formula = occ.formula,
det.formula = det.formula,
data = data.list,
inits = inits.list,
n.batch = n.batch,
batch.length = batch.length,
accept.rate = 0.43,
priors = prior.list,
cov.model = "exponential",
tuning = tuning.list,
n.omp.threads = 1,
verbose = TRUE,
NNGP = TRUE,
n.neighbors = 5,
search.type = 'cb',
n.report = 10,
n.burn = 100,
n.thin = 1,
n.chains = 1))
})
# Check waicOcc -----------------------
test_that("waicOCC works for spMsPGOcc", {
# as.vector gets rid of names
waic.out <- as.vector(waicOcc(out))
expect_equal(length(waic.out), 3)
expect_equal(waic.out[3], -2 * (waic.out[1] - waic.out[2]))
})
test_that("fitted works for spMsPGOcc", {
fitted.out <- fitted(out)
expect_equal(length(fitted.out), 2)
expect_equal(class(fitted.out$y.rep.samples), "array")
expect_equal(class(fitted.out$p.samples), "array")
expect_equal(dim(fitted.out$y.rep.samples), dim(fitted.out$p.samples))
})
test_that("predict works for spMsPGOcc", {
n.post.samples <- out$n.post * out$n.chains
X.0 <- cbind(X.0, X.re.0)
colnames(X.0) <- c('int', 'occ.factor.1', 'occ.factor.2')
pred.out <- predict(out, X.0, coords.0, verbose = FALSE)
expect_type(pred.out, "list")
expect_equal(dim(pred.out$psi.0.samples), c(n.post.samples, N, nrow(X.0)))
expect_equal(dim(pred.out$z.0.samples), c(n.post.samples, N, nrow(X.0)))
})
test_that("detection prediction works", {
J.str <- 100
X.p.0 <- matrix(1, nrow = J.str, ncol = p.det)
pred.out <- predict(out, X.p.0, type = 'detection')
expect_type(pred.out, 'list')
expect_equal(dim(pred.out$p.0.samples), c(out$n.post * out$n.chains, N, J.str))
})
test_that("posterior predictive checks work for msPGOcc", {
n.post.samples <- out$n.post * out$n.chains
J.fit <- nrow(X)
ppc.out <- ppcOcc(out, 'chi-square', 2)
expect_type(ppc.out, "list")
expect_equal(dim(ppc.out$fit.y), c(n.post.samples, N))
expect_equal(dim(ppc.out$fit.y.rep), c(n.post.samples, N))
expect_equal(dim(ppc.out$fit.y.group.quants), c(5, N, max(n.rep)))
expect_equal(dim(ppc.out$fit.y.rep.group.quants), c(5, N, max(n.rep)))
ppc.out <- ppcOcc(out, 'chi-square', 1)
expect_type(ppc.out, "list")
expect_equal(dim(ppc.out$fit.y), c(n.post.samples, N))
expect_equal(dim(ppc.out$fit.y.rep), c(n.post.samples, N))
expect_equal(dim(ppc.out$fit.y.group.quants), c(5, N, J.fit))
expect_equal(dim(ppc.out$fit.y.rep.group.quants), c(5, N, J.fit))
ppc.out <- ppcOcc(out, 'freeman-tukey', 1)
expect_type(ppc.out, "list")
expect_equal(dim(ppc.out$fit.y), c(n.post.samples, N))
expect_equal(dim(ppc.out$fit.y.rep), c(n.post.samples, N))
expect_equal(dim(ppc.out$fit.y.group.quants), c(5, N, J.fit))
expect_equal(dim(ppc.out$fit.y.rep.group.quants), c(5, N, J.fit))
ppc.out <- ppcOcc(out, 'freeman-tukey', 2)
expect_type(ppc.out, "list")
expect_equal(dim(ppc.out$fit.y), c(n.post.samples, N))
expect_equal(dim(ppc.out$fit.y.rep), c(n.post.samples, N))
expect_equal(dim(ppc.out$fit.y.group.quants), c(5, N, max(n.rep)))
expect_equal(dim(ppc.out$fit.y.rep.group.quants), c(5, N, max(n.rep)))
})
# Occurrence REs + covariates ---------------------------------------------
J.x <- 8
J.y <- 8
J <- J.x * J.y
n.rep<- sample(2:4, size = J, replace = TRUE)
N <- 6
# Community-level covariate effects
# Occurrence
beta.mean <- c(0.2, 1.2, -1.9)
p.occ <- length(beta.mean)
tau.sq.beta <- c(0.6, 1.5, 2.3)
# Detection
alpha.mean <- c(0)
tau.sq.alpha <- c(1)
p.det <- length(alpha.mean)
# Random effects
psi.RE <- list(levels = c(20, 10),
sigma.sq.psi = c(1.3, 3.4))
p.RE <- list()
# Draw species-level effects from community means.
beta <- matrix(NA, nrow = N, ncol = p.occ)
alpha <- matrix(NA, nrow = N, ncol = p.det)
for (i in 1:p.occ) {
beta[, i] <- rnorm(N, beta.mean[i], sqrt(tau.sq.beta[i]))
}
for (i in 1:p.det) {
alpha[, i] <- rnorm(N, alpha.mean[i], sqrt(tau.sq.alpha[i]))
}
alpha.true <- alpha
phi <- rep(3 / .7, N)
sigma.sq <- rep(2, N)
nu <- rep(2, N)
dat <- simMsOcc(J.x = J.x, J.y = J.y, n.rep = n.rep, N = N, beta = beta, alpha = alpha,
psi.RE = psi.RE, p.RE = p.RE, sp = TRUE, sigma.sq = sigma.sq,
phi = phi, nu = nu, cov.model = 'matern')
pred.indx <- sample(1:J, round(J * .25), replace = FALSE)
y <- dat$y[, -pred.indx, , drop = FALSE]
# Occupancy covariates
X <- dat$X[-pred.indx, , drop = FALSE]
X.re <- dat$X.re[-pred.indx, , drop = FALSE]
coords <- as.matrix(dat$coords[-pred.indx, , drop = FALSE])
# Prediction covariates
X.0 <- dat$X[pred.indx, , drop = FALSE]
X.re.0 <- dat$X.re[pred.indx, , drop = FALSE]
coords.0 <- as.matrix(dat$coords[pred.indx, , drop = FALSE])
# Detection covariates
X.p <- dat$X.p[-pred.indx, , , drop = FALSE]
occ.covs <- cbind(X, X.re)
colnames(occ.covs) <- c('int', 'occ.cov.1', 'occ.cov.2', 'occ.factor.1', 'occ.factor.2')
data.list <- list(y = y, coords = coords, occ.covs = occ.covs)
# Priors
prior.list <- list(beta.comm.normal = list(mean = 0, var = 2.72),
alpha.comm.normal = list(mean = 0, var = 2.72),
tau.sq.beta.ig = list(a = 0.1, b = 0.1),
tau.sq.alpha.ig = list(a = 0.1, b = 0.1),
nu.unif = list(0.5, 2.5))
# Starting values
inits.list <- list(alpha.comm = 0,
beta.comm = 0,
beta = 0,
alpha = 0,
tau.sq.beta = 1,
tau.sq.alpha = 1,
z = apply(y, c(1, 2), max, na.rm = TRUE))
# Tuning
tuning.list <- list(phi = 1, nu = 0.25)
batch.length <- 25
n.batch <- 10
n.report <- 100
occ.formula <- ~ occ.cov.1 + occ.cov.2 + (1 | occ.factor.1) + (1 | occ.factor.2)
det.formula <- ~ 1
out <- spMsPGOcc(occ.formula = occ.formula,
det.formula = det.formula,
data = data.list,
inits = inits.list,
n.batch = n.batch,
batch.length = batch.length,
accept.rate = 0.43,
priors = prior.list,
cov.model = "matern",
tuning = tuning.list,
n.omp.threads = 1,
verbose = FALSE,
NNGP = TRUE,
n.neighbors = 5,
search.type = 'cb',
n.report = 10,
n.burn = 100,
n.thin = 2,
n.chains = 2,
k.fold = 2,
k.fold.threads = 1)
# To make sure it worked --------------
test_that("out is of class spMsPGOcc", {
expect_s3_class(out, "spMsPGOcc")
})
# Check cross-validation --------------
test_that("cross-validation works", {
expect_equal(length(out$k.fold.deviance), N)
expect_type(out$k.fold.deviance, "double")
expect_equal(sum(out$k.fold.deviance < 0), 0)
})
# Check random effects ----------------
test_that("random effects are correct", {
expect_equal(out$pRE, FALSE)
expect_equal(out$psiRE, TRUE)
})
# Check RE error ----------------------
test_that("random effect gives error when non-numeric", {
data.list$occ.covs <- as.data.frame(data.list$occ.covs)
data.list$occ.covs$occ.factor.1 <- factor(data.list$occ.covs$occ.factor.1)
# data.list$det.covs$det.factor.1 <- factor(data.list$det.covs$det.factor.1)
expect_error(out <- spMsPGOcc(occ.formula = occ.formula,
det.formula = det.formula,
data = data.list,
inits = inits.list,
n.batch = n.batch,
batch.length = batch.length,
accept.rate = 0.43,
priors = prior.list,
cov.model = "matern",
tuning = tuning.list,
n.omp.threads = 1,
verbose = FALSE,
NNGP = TRUE,
n.neighbors = 5,
search.type = 'cb',
n.report = 10,
n.burn = 100,
n.thin = 2,
n.chains = 1))
data.list$occ.covs$occ.factor.1 <- as.character(factor(data.list$occ.covs$occ.factor.1))
# data.list$det.covs$det.factor.1 <- as.character(factor(data.list$det.covs$det.factor.1))
expect_error(out <- spMsPGOcc(occ.formula = occ.formula,
det.formula = det.formula,
data = data.list,
inits = inits.list,
n.batch = n.batch,
batch.length = batch.length,
accept.rate = 0.43,
priors = prior.list,
cov.model = "matern",
tuning = tuning.list,
n.omp.threads = 1,
verbose = FALSE,
NNGP = TRUE,
n.neighbors = 5,
search.type = 'cb',
n.report = 10,
n.burn = 100,
n.thin = 2,
n.chains = 1))
})
# Check RE levels ---------------------
test_that("random effect levels are correct", {
expect_equal(sort(unique(unlist(out$re.level.names))),
sort(unique(c(X.re))))
})
# Check output data output is correct -
test_that("out$y == y", {
expect_equal(out$y, y)
})
# Check missing values ----------------
test_that("missing value error handling works", {
tmp.data <- data.list
tmp.data$occ.covs[3, ] <- NA
expect_error(spMsPGOcc(occ.formula = occ.formula,
det.formula = det.formula,
data = tmp.data,
n.batch = 40,
batch.length = batch.length,
cov.model = "exponential",
tuning = tuning.list,
NNGP = TRUE,
verbose = FALSE,
n.neighbors = 5,
search.type = 'cb',
n.report = 10,
n.burn = 100,
n.chains = 1))
# tmp.data <- data.list
# tmp.data$det.covs[[1]][1] <- NA
# expect_error(spMsPGOcc(occ.formula = occ.formula,
# det.formula = det.formula,
# data = tmp.data,
# n.batch = 40,
# batch.length = batch.length,
# cov.model = "exponential",
# tuning = tuning.list,
# NNGP = TRUE,
# verbose = FALSE,
# n.neighbors = 5,
# search.type = 'cb',
# n.report = 10,
# n.burn = 100,
# n.chains = 1))
# tmp.data <- data.list
# tmp.data$y[, 1, 1] <- NA
# out <- spMsPGOcc(occ.formula = occ.formula,
# det.formula = det.formula,
# data = tmp.data,
# n.batch = 40,
# batch.length = batch.length,
# cov.model = "exponential",
# tuning = tuning.list,
# NNGP = TRUE,
# verbose = FALSE,
# n.neighbors = 5,
# search.type = 'cb',
# n.report = 10,
# n.burn = 100,
# n.chains = 1)
# expect_s3_class(out, "spMsPGOcc")
})
# Check default values ----------------
test_that("default priors, inits, burn, thin work", {
out <- spMsPGOcc(occ.formula = occ.formula,
det.formula = det.formula,
data = data.list,
n.batch = n.batch,
batch.length = batch.length,
accept.rate = 0.43,
cov.model = "exponential",
tuning = tuning.list,
NNGP = TRUE,
verbose = FALSE,
n.neighbors = 5,
n.chains = 1)
expect_s3_class(out, "spMsPGOcc")
})
test_that("all correlation functions work", {
out <- spMsPGOcc(occ.formula = occ.formula,
det.formula = det.formula,
data = data.list,
inits = inits.list,
n.batch = n.batch,
batch.length = batch.length,
accept.rate = 0.43,
priors = prior.list,
cov.model = "gaussian",
tuning = list(phi = 0.3),
n.omp.threads = 1,
verbose = FALSE,
NNGP = TRUE,
n.neighbors = 5,
search.type = 'cb',
n.report = 10,
n.burn = 100,
n.thin = 1,
n.chains = 1)
expect_s3_class(out, "spMsPGOcc")
out <- spMsPGOcc(occ.formula = occ.formula,
det.formula = det.formula,
data = data.list,
inits = inits.list,
n.batch = n.batch,
batch.length = batch.length,
accept.rate = 0.43,
priors = prior.list,
cov.model = "spherical",
tuning = list(phi = 0.3),
n.omp.threads = 1,
verbose = FALSE,
NNGP = TRUE,
n.neighbors = 5,
search.type = 'cb',
n.report = 10,
n.burn = 100,
n.thin = 1,
n.chains = 1)
expect_s3_class(out, "spMsPGOcc")
out <- spMsPGOcc(occ.formula = occ.formula,
det.formula = det.formula,
data = data.list,
inits = inits.list,
n.batch = n.batch,
batch.length = batch.length,
accept.rate = 0.43,
priors = list(nu.unif = list(0.4, 3)),
cov.model = "matern",
tuning = list(phi = 0.3, nu = 0.2),
n.omp.threads = 1,
verbose = FALSE,
NNGP = TRUE,
n.neighbors = 5,
search.type = 'cb',
n.report = 10,
n.burn = 100,
n.thin = 1,
n.chains = 1)
expect_s3_class(out, "spMsPGOcc")
})
test_that("verbose prints to the screen", {
expect_output(spMsPGOcc(occ.formula = occ.formula,
det.formula = det.formula,
data = data.list,
inits = inits.list,
n.batch = n.batch,
batch.length = batch.length,
accept.rate = 0.43,
priors = prior.list,
cov.model = "exponential",
tuning = tuning.list,
n.omp.threads = 1,
verbose = TRUE,
NNGP = TRUE,
n.neighbors = 5,
search.type = 'cb',
n.report = 10,
n.burn = 100,
n.thin = 1,
n.chains = 1))
})
# Check waicOcc -----------------------
test_that("waicOCC works for spMsPGOcc", {
# as.vector gets rid of names
waic.out <- as.vector(waicOcc(out))
expect_equal(length(waic.out), 3)
expect_equal(waic.out[3], -2 * (waic.out[1] - waic.out[2]))
})
test_that("fitted works for spMsPGOcc", {
fitted.out <- fitted(out)
expect_equal(length(fitted.out), 2)
expect_equal(class(fitted.out$y.rep.samples), "array")
expect_equal(class(fitted.out$p.samples), "array")
expect_equal(dim(fitted.out$y.rep.samples), dim(fitted.out$p.samples))
})
test_that("predict works for spMsPGOcc", {
n.post.samples <- out$n.post * out$n.chains
X.0 <- cbind(X.0, X.re.0)
colnames(X.0) <- c('int', 'occ.cov.1', 'occ.cov.2', 'occ.factor.1', 'occ.factor.2')
pred.out <- predict(out, X.0, coords.0, verbose = FALSE)
expect_type(pred.out, "list")
expect_equal(dim(pred.out$psi.0.samples), c(n.post.samples, N, nrow(X.0)))
expect_equal(dim(pred.out$z.0.samples), c(n.post.samples, N, nrow(X.0)))
})
test_that("detection prediction works", {
J.str <- 100
X.p.0 <- matrix(1, nrow = J.str, ncol = p.det)
pred.out <- predict(out, X.p.0, type = 'detection')
expect_type(pred.out, 'list')
expect_equal(dim(pred.out$p.0.samples), c(out$n.post * out$n.chains, N, J.str))
})
test_that("posterior predictive checks work for msPGOcc", {
n.post.samples <- out$n.post * out$n.chains
J.fit <- nrow(X)
ppc.out <- ppcOcc(out, 'chi-square', 2)
expect_type(ppc.out, "list")
expect_equal(dim(ppc.out$fit.y), c(n.post.samples, N))
expect_equal(dim(ppc.out$fit.y.rep), c(n.post.samples, N))
expect_equal(dim(ppc.out$fit.y.group.quants), c(5, N, max(n.rep)))
expect_equal(dim(ppc.out$fit.y.rep.group.quants), c(5, N, max(n.rep)))
ppc.out <- ppcOcc(out, 'chi-square', 1)
expect_type(ppc.out, "list")
expect_equal(dim(ppc.out$fit.y), c(n.post.samples, N))
expect_equal(dim(ppc.out$fit.y.rep), c(n.post.samples, N))
expect_equal(dim(ppc.out$fit.y.group.quants), c(5, N, J.fit))
expect_equal(dim(ppc.out$fit.y.rep.group.quants), c(5, N, J.fit))
ppc.out <- ppcOcc(out, 'freeman-tukey', 1)
expect_type(ppc.out, "list")
expect_equal(dim(ppc.out$fit.y), c(n.post.samples, N))
expect_equal(dim(ppc.out$fit.y.rep), c(n.post.samples, N))
expect_equal(dim(ppc.out$fit.y.group.quants), c(5, N, J.fit))
expect_equal(dim(ppc.out$fit.y.rep.group.quants), c(5, N, J.fit))
ppc.out <- ppcOcc(out, 'freeman-tukey', 2)
expect_type(ppc.out, "list")
expect_equal(dim(ppc.out$fit.y), c(n.post.samples, N))
expect_equal(dim(ppc.out$fit.y.rep), c(n.post.samples, N))
expect_equal(dim(ppc.out$fit.y.group.quants), c(5, N, max(n.rep)))
expect_equal(dim(ppc.out$fit.y.rep.group.quants), c(5, N, max(n.rep)))
})
# Occurrence REs + covariates in everything -------------------------------
J.x <- 8
J.y <- 8
J <- J.x * J.y
n.rep<- sample(2:4, size = J, replace = TRUE)
N <- 6
# Community-level covariate effects
# Occurrence
beta.mean <- c(0.2, 1.2, -1.9)
p.occ <- length(beta.mean)
tau.sq.beta <- c(0.6, 1.5, 2.3)
# Detection
alpha.mean <- c(0, 1.2)
tau.sq.alpha <- c(1, 2.3)
p.det <- length(alpha.mean)
# Random effects
psi.RE <- list(levels = c(20, 10),
sigma.sq.psi = c(1.3, 3.4))
p.RE <- list()
# Draw species-level effects from community means.
beta <- matrix(NA, nrow = N, ncol = p.occ)
alpha <- matrix(NA, nrow = N, ncol = p.det)
for (i in 1:p.occ) {
beta[, i] <- rnorm(N, beta.mean[i], sqrt(tau.sq.beta[i]))
}
for (i in 1:p.det) {
alpha[, i] <- rnorm(N, alpha.mean[i], sqrt(tau.sq.alpha[i]))
}
alpha.true <- alpha
phi <- rep(3 / .7, N)
sigma.sq <- rep(2, N)
nu <- rep(2, N)
dat <- simMsOcc(J.x = J.x, J.y = J.y, n.rep = n.rep, N = N, beta = beta, alpha = alpha,
psi.RE = psi.RE, p.RE = p.RE, sp = TRUE, sigma.sq = sigma.sq,
phi = phi, nu = nu, cov.model = 'matern')
pred.indx <- sample(1:J, round(J * .25), replace = FALSE)
y <- dat$y[, -pred.indx, , drop = FALSE]
# Occupancy covariates
X <- dat$X[-pred.indx, , drop = FALSE]
X.re <- dat$X.re[-pred.indx, , drop = FALSE]
coords <- as.matrix(dat$coords[-pred.indx, , drop = FALSE])
# Prediction covariates
X.0 <- dat$X[pred.indx, , drop = FALSE]
X.re.0 <- dat$X.re[pred.indx, , drop = FALSE]
coords.0 <- as.matrix(dat$coords[pred.indx, , drop = FALSE])
# Detection covariates
X.p <- dat$X.p[-pred.indx, , , drop = FALSE]
occ.covs <- cbind(X, X.re)
colnames(occ.covs) <- c('int', 'occ.cov.1', 'occ.cov.2', 'occ.factor.1', 'occ.factor.2')
det.covs <- list(det.cov.1 = X.p[, , 2])
data.list <- list(y = y, coords = coords, occ.covs = occ.covs, det.covs = det.covs)
# Priors
prior.list <- list(beta.comm.normal = list(mean = 0, var = 2.72),
alpha.comm.normal = list(mean = 0, var = 2.72),
tau.sq.beta.ig = list(a = 0.1, b = 0.1),
tau.sq.alpha.ig = list(a = 0.1, b = 0.1),
nu.unif = list(0.5, 2.5))
# Starting values
inits.list <- list(alpha.comm = 0,
beta.comm = 0,
beta = 0,
alpha = 0,
tau.sq.beta = 1,
tau.sq.alpha = 1,
z = apply(y, c(1, 2), max, na.rm = TRUE))
# Tuning
tuning.list <- list(phi = 1, nu = 0.25)
batch.length <- 25
n.batch <- 10
n.report <- 100
occ.formula <- ~ occ.cov.1 + occ.cov.2 + (1 | occ.factor.1) + (1 | occ.factor.2)
det.formula <- ~ det.cov.1
out <- spMsPGOcc(occ.formula = occ.formula,
det.formula = det.formula,
data = data.list,
inits = inits.list,
n.batch = n.batch,
batch.length = batch.length,
accept.rate = 0.43,
priors = prior.list,
cov.model = "matern",
tuning = tuning.list,
n.omp.threads = 1,
verbose = FALSE,
NNGP = TRUE,
n.neighbors = 5,
search.type = 'cb',
n.report = 10,
n.burn = 100,
n.thin = 2,
n.chains = 2,
k.fold = 2,
k.fold.threads = 1)
# To make sure it worked --------------
test_that("out is of class spMsPGOcc", {
expect_s3_class(out, "spMsPGOcc")
})
# Check cross-validation --------------
test_that("cross-validation works", {
expect_equal(length(out$k.fold.deviance), N)
expect_type(out$k.fold.deviance, "double")
expect_equal(sum(out$k.fold.deviance < 0), 0)
})
# Check random effects ----------------
test_that("random effects are correct", {
expect_equal(out$pRE, FALSE)
expect_equal(out$psiRE, TRUE)
})
# Check RE error ----------------------
test_that("random effect gives error when non-numeric", {
data.list$occ.covs <- as.data.frame(data.list$occ.covs)
data.list$occ.covs$occ.factor.1 <- factor(data.list$occ.covs$occ.factor.1)
# data.list$det.covs$det.factor.1 <- factor(data.list$det.covs$det.factor.1)
expect_error(out <- spMsPGOcc(occ.formula = occ.formula,
det.formula = det.formula,
data = data.list,
inits = inits.list,
n.batch = n.batch,
batch.length = batch.length,
accept.rate = 0.43,
priors = prior.list,
cov.model = "matern",
tuning = tuning.list,
n.omp.threads = 1,
verbose = FALSE,
NNGP = TRUE,
n.neighbors = 5,
search.type = 'cb',
n.report = 10,
n.burn = 100,
n.thin = 2,
n.chains = 1))
data.list$occ.covs$occ.factor.1 <- as.character(factor(data.list$occ.covs$occ.factor.1))
# data.list$det.covs$det.factor.1 <- as.character(factor(data.list$det.covs$det.factor.1))
expect_error(out <- spMsPGOcc(occ.formula = occ.formula,
det.formula = det.formula,
data = data.list,
inits = inits.list,
n.batch = n.batch,
batch.length = batch.length,
accept.rate = 0.43,
priors = prior.list,
cov.model = "matern",
tuning = tuning.list,
n.omp.threads = 1,
verbose = FALSE,
NNGP = TRUE,
n.neighbors = 5,
search.type = 'cb',
n.report = 10,
n.burn = 100,
n.thin = 2,
n.chains = 1))
})
# Check RE levels ---------------------
test_that("random effect levels are correct", {
expect_equal(sort(unique(unlist(out$re.level.names))),
sort(unique(c(X.re))))
})
# Check output data output is correct -
test_that("out$y == y", {
expect_equal(out$y, y)
})
# Check missing values ----------------
test_that("missing value error handling works", {
tmp.data <- data.list
tmp.data$occ.covs[3, ] <- NA
expect_error(spMsPGOcc(occ.formula = occ.formula,
det.formula = det.formula,
data = tmp.data,
n.batch = 40,
batch.length = batch.length,
cov.model = "exponential",
tuning = tuning.list,
NNGP = TRUE,
verbose = FALSE,
n.neighbors = 5,
search.type = 'cb',
n.report = 10,
n.burn = 100,
n.chains = 1))
tmp.data <- data.list
tmp.data$y[1, 1, 1] <- 1
tmp.data$det.covs[[1]][1] <- NA
expect_error(spMsPGOcc(occ.formula = occ.formula,
det.formula = det.formula,
data = tmp.data,
n.batch = 40,
batch.length = batch.length,
cov.model = "exponential",
tuning = tuning.list,
NNGP = TRUE,
verbose = FALSE,
n.neighbors = 5,
search.type = 'cb',
n.report = 10,
n.burn = 100,
n.chains = 1))
tmp.data <- data.list
tmp.data$y[, 1, 1] <- NA
out <- spMsPGOcc(occ.formula = occ.formula,
det.formula = det.formula,
data = tmp.data,
n.batch = 40,
batch.length = batch.length,
cov.model = "exponential",
tuning = tuning.list,
NNGP = TRUE,
verbose = FALSE,
n.neighbors = 5,
search.type = 'cb',
n.report = 10,
n.burn = 100,
n.chains = 1)
expect_s3_class(out, "spMsPGOcc")
})
# Check default values ----------------
test_that("default priors, inits, burn, thin work", {
out <- spMsPGOcc(occ.formula = occ.formula,
det.formula = det.formula,
data = data.list,
n.batch = n.batch,
batch.length = batch.length,
accept.rate = 0.43,
cov.model = "exponential",
tuning = tuning.list,
NNGP = TRUE,
verbose = FALSE,
n.neighbors = 5,
n.chains = 1)
expect_s3_class(out, "spMsPGOcc")
})
test_that("all correlation functions work", {
out <- spMsPGOcc(occ.formula = occ.formula,
det.formula = det.formula,
data = data.list,
inits = inits.list,
n.batch = n.batch,
batch.length = batch.length,
accept.rate = 0.43,
priors = prior.list,
cov.model = "gaussian",
tuning = list(phi = 0.3),
n.omp.threads = 1,
verbose = FALSE,
NNGP = TRUE,
n.neighbors = 5,
search.type = 'cb',
n.report = 10,
n.burn = 100,
n.thin = 1,
n.chains = 1)
expect_s3_class(out, "spMsPGOcc")
out <- spMsPGOcc(occ.formula = occ.formula,
det.formula = det.formula,
data = data.list,
inits = inits.list,
n.batch = n.batch,
batch.length = batch.length,
accept.rate = 0.43,
priors = prior.list,
cov.model = "spherical",
tuning = list(phi = 0.3),
n.omp.threads = 1,
verbose = FALSE,
NNGP = TRUE,
n.neighbors = 5,
search.type = 'cb',
n.report = 10,
n.burn = 100,
n.thin = 1,
n.chains = 1)
expect_s3_class(out, "spMsPGOcc")
out <- spMsPGOcc(occ.formula = occ.formula,
det.formula = det.formula,
data = data.list,
inits = inits.list,
n.batch = n.batch,
batch.length = batch.length,
accept.rate = 0.43,
priors = list(nu.unif = list(0.4, 3)),
cov.model = "matern",
tuning = list(phi = 0.3, nu = 0.2),
n.omp.threads = 1,
verbose = FALSE,
NNGP = TRUE,
n.neighbors = 5,
search.type = 'cb',
n.report = 10,
n.burn = 100,
n.thin = 1,
n.chains = 1)
expect_s3_class(out, "spMsPGOcc")
})
test_that("verbose prints to the screen", {
expect_output(spMsPGOcc(occ.formula = occ.formula,
det.formula = det.formula,
data = data.list,
inits = inits.list,
n.batch = n.batch,
batch.length = batch.length,
accept.rate = 0.43,
priors = prior.list,
cov.model = "exponential",
tuning = tuning.list,
n.omp.threads = 1,
verbose = TRUE,
NNGP = TRUE,
n.neighbors = 5,
search.type = 'cb',
n.report = 10,
n.burn = 100,
n.thin = 1,
n.chains = 1))
})
# Check waicOcc -----------------------
test_that("waicOCC works for spMsPGOcc", {
# as.vector gets rid of names
waic.out <- as.vector(waicOcc(out))
expect_equal(length(waic.out), 3)
expect_equal(waic.out[3], -2 * (waic.out[1] - waic.out[2]))
})
test_that("fitted works for spMsPGOcc", {
fitted.out <- fitted(out)
expect_equal(length(fitted.out), 2)
expect_equal(class(fitted.out$y.rep.samples), "array")
expect_equal(class(fitted.out$p.samples), "array")
expect_equal(dim(fitted.out$y.rep.samples), dim(fitted.out$p.samples))
})
test_that("predict works for spMsPGOcc", {
n.post.samples <- out$n.post * out$n.chains
X.0 <- cbind(X.0, X.re.0)
colnames(X.0) <- c('int', 'occ.cov.1', 'occ.cov.2', 'occ.factor.1', 'occ.factor.2')
pred.out <- predict(out, X.0, coords.0, verbose = FALSE)
expect_type(pred.out, "list")
expect_equal(dim(pred.out$psi.0.samples), c(n.post.samples, N, nrow(X.0)))
expect_equal(dim(pred.out$z.0.samples), c(n.post.samples, N, nrow(X.0)))
})
test_that("detection prediction works", {
X.p.0 <- dat$X.p[, 1, ]
pred.out <- predict(out, X.p.0, type = 'detection')
expect_type(pred.out, 'list')
expect_equal(dim(pred.out$p.0.samples), c(out$n.post * out$n.chains, N, J))
})
test_that("posterior predictive checks work for msPGOcc", {
n.post.samples <- out$n.post * out$n.chains
J.fit <- nrow(X)
ppc.out <- ppcOcc(out, 'chi-square', 2)
expect_type(ppc.out, "list")
expect_equal(dim(ppc.out$fit.y), c(n.post.samples, N))
expect_equal(dim(ppc.out$fit.y.rep), c(n.post.samples, N))
expect_equal(dim(ppc.out$fit.y.group.quants), c(5, N, max(n.rep)))
expect_equal(dim(ppc.out$fit.y.rep.group.quants), c(5, N, max(n.rep)))
ppc.out <- ppcOcc(out, 'chi-square', 1)
expect_type(ppc.out, "list")
expect_equal(dim(ppc.out$fit.y), c(n.post.samples, N))
expect_equal(dim(ppc.out$fit.y.rep), c(n.post.samples, N))
expect_equal(dim(ppc.out$fit.y.group.quants), c(5, N, J.fit))
expect_equal(dim(ppc.out$fit.y.rep.group.quants), c(5, N, J.fit))
ppc.out <- ppcOcc(out, 'freeman-tukey', 1)
expect_type(ppc.out, "list")
expect_equal(dim(ppc.out$fit.y), c(n.post.samples, N))
expect_equal(dim(ppc.out$fit.y.rep), c(n.post.samples, N))
expect_equal(dim(ppc.out$fit.y.group.quants), c(5, N, J.fit))
expect_equal(dim(ppc.out$fit.y.rep.group.quants), c(5, N, J.fit))
ppc.out <- ppcOcc(out, 'freeman-tukey', 2)
expect_type(ppc.out, "list")
expect_equal(dim(ppc.out$fit.y), c(n.post.samples, N))
expect_equal(dim(ppc.out$fit.y.rep), c(n.post.samples, N))
expect_equal(dim(ppc.out$fit.y.group.quants), c(5, N, max(n.rep)))
expect_equal(dim(ppc.out$fit.y.rep.group.quants), c(5, N, max(n.rep)))
})
# Random intercept on detection -------------------------------------------
J.x <- 8
J.y <- 8
J <- J.x * J.y
n.rep<- sample(2:4, size = J, replace = TRUE)
N <- 6
# Community-level covariate effects
# Occurrence
beta.mean <- c(0.2)
p.occ <- length(beta.mean)
tau.sq.beta <- c(0.6)
# Detection
alpha.mean <- c(0)
tau.sq.alpha <- c(1)
p.det <- length(alpha.mean)
# Random effects
psi.RE <- list()
p.RE <- list(levels = c(50),
sigma.sq.p = c(2.50))
# Draw species-level effects from community means.
beta <- matrix(NA, nrow = N, ncol = p.occ)
alpha <- matrix(NA, nrow = N, ncol = p.det)
for (i in 1:p.occ) {
beta[, i] <- rnorm(N, beta.mean[i], sqrt(tau.sq.beta[i]))
}
for (i in 1:p.det) {
alpha[, i] <- rnorm(N, alpha.mean[i], sqrt(tau.sq.alpha[i]))
}
alpha.true <- alpha
phi <- rep(3 / .7, N)
sigma.sq <- rep(2, N)
nu <- rep(2, N)
dat <- simMsOcc(J.x = J.x, J.y = J.y, n.rep = n.rep, N = N, beta = beta, alpha = alpha,
psi.RE = psi.RE, p.RE = p.RE, sp = TRUE, sigma.sq = sigma.sq,
phi = phi, nu = nu, cov.model = 'matern')
pred.indx <- sample(1:J, round(J * .25), replace = FALSE)
y <- dat$y[, -pred.indx, , drop = FALSE]
# Occupancy covariates
X <- dat$X[-pred.indx, , drop = FALSE]
# X.re <- dat$X.re[-pred.indx, , drop = FALSE]
coords <- as.matrix(dat$coords[-pred.indx, , drop = FALSE])
# Prediction covariates
X.0 <- dat$X[pred.indx, , drop = FALSE]
# X.re.0 <- dat$X.re[pred.indx, , drop = FALSE]
coords.0 <- as.matrix(dat$coords[pred.indx, , drop = FALSE])
# Detection covariates
X.p <- dat$X.p[-pred.indx, , , drop = FALSE]
X.p.re <- dat$X.p.re[-pred.indx, , , drop = FALSE]
occ.covs <- cbind(X)
colnames(occ.covs) <- c('int')
det.covs <- list(det.factor.1 = X.p.re[, , 1])
data.list <- list(y = y, coords = coords, occ.covs = occ.covs, det.covs = det.covs)
# Priors
prior.list <- list(beta.comm.normal = list(mean = 0, var = 2.72),
alpha.comm.normal = list(mean = 0, var = 2.72),
tau.sq.beta.ig = list(a = 0.1, b = 0.1),
tau.sq.alpha.ig = list(a = 0.1, b = 0.1),
nu.unif = list(0.5, 2.5))
# Starting values
inits.list <- list(alpha.comm = 0,
beta.comm = 0,
beta = 0,
alpha = 0,
tau.sq.beta = 1,
tau.sq.alpha = 1,
z = apply(y, c(1, 2), max, na.rm = TRUE))
# Tuning
tuning.list <- list(phi = 1, nu = 0.25)
batch.length <- 25
n.batch <- 10
n.report <- 100
occ.formula <- ~ 1
det.formula <- ~ (1 | det.factor.1)
out <- spMsPGOcc(occ.formula = occ.formula,
det.formula = det.formula,
data = data.list,
inits = inits.list,
n.batch = n.batch,
batch.length = batch.length,
accept.rate = 0.43,
priors = prior.list,
cov.model = "matern",
tuning = tuning.list,
n.omp.threads = 1,
verbose = FALSE,
NNGP = TRUE,
n.neighbors = 5,
search.type = 'cb',
n.report = 10,
n.burn = 100,
n.thin = 2,
n.chains = 2,
k.fold = 2,
k.fold.threads = 1)
# To make sure it worked --------------
test_that("out is of class spMsPGOcc", {
expect_s3_class(out, "spMsPGOcc")
})
# Check cross-validation --------------
test_that("cross-validation works", {
expect_equal(length(out$k.fold.deviance), N)
expect_type(out$k.fold.deviance, "double")
expect_equal(sum(out$k.fold.deviance < 0), 0)
})
# Check random effects ----------------
test_that("random effects are correct", {
expect_equal(out$pRE, TRUE)
expect_equal(out$psiRE, FALSE)
})
# Check RE error ----------------------
test_that("random effect gives error when non-numeric", {
# data.list$occ.covs <- as.data.frame(data.list$occ.covs)
# data.list$occ.covs$occ.factor.1 <- factor(data.list$occ.covs$occ.factor.1)
data.list$det.covs$det.factor.1 <- factor(data.list$det.covs$det.factor.1)
expect_error(out <- spMsPGOcc(occ.formula = occ.formula,
det.formula = det.formula,
data = data.list,
inits = inits.list,
n.batch = n.batch,
batch.length = batch.length,
accept.rate = 0.43,
priors = prior.list,
cov.model = "matern",
tuning = tuning.list,
n.omp.threads = 1,
verbose = FALSE,
NNGP = TRUE,
n.neighbors = 5,
search.type = 'cb',
n.report = 10,
n.burn = 100,
n.thin = 2,
n.chains = 1))
# data.list$occ.covs$occ.factor.1 <- as.character(factor(data.list$occ.covs$occ.factor.1))
data.list$det.covs$det.factor.1 <- as.character(factor(data.list$det.covs$det.factor.1))
expect_error(out <- spMsPGOcc(occ.formula = occ.formula,
det.formula = det.formula,
data = data.list,
inits = inits.list,
n.batch = n.batch,
batch.length = batch.length,
accept.rate = 0.43,
priors = prior.list,
cov.model = "matern",
tuning = tuning.list,
n.omp.threads = 1,
verbose = FALSE,
NNGP = TRUE,
n.neighbors = 5,
search.type = 'cb',
n.report = 10,
n.burn = 100,
n.thin = 2,
n.chains = 1))
})
# Check RE levels ---------------------
test_that("random effect levels are correct", {
expect_equal(sort(unique(unlist(out$p.re.level.names))),
sort(unique(c(X.p.re))))
})
# Check output data output is correct -
test_that("out$y == y", {
expect_equal(out$y, y)
})
# Check missing values ----------------
test_that("missing value error handling works", {
# tmp.data <- data.list
# tmp.data$occ.covs[3, ] <- NA
# expect_error(spMsPGOcc(occ.formula = occ.formula,
# det.formula = det.formula,
# data = tmp.data,
# n.batch = 40,
# batch.length = batch.length,
# cov.model = "exponential",
# tuning = tuning.list,
# NNGP = TRUE,
# verbose = FALSE,
# n.neighbors = 5,
# search.type = 'cb',
# n.report = 10,
# n.burn = 100,
# n.chains = 1))
tmp.data <- data.list
tmp.data$y[1, 1, 1] <- 1
tmp.data$det.covs[[1]][1] <- NA
expect_error(spMsPGOcc(occ.formula = occ.formula,
det.formula = det.formula,
data = tmp.data,
n.batch = 40,
batch.length = batch.length,
cov.model = "exponential",
tuning = tuning.list,
NNGP = TRUE,
verbose = FALSE,
n.neighbors = 5,
search.type = 'cb',
n.report = 10,
n.burn = 100,
n.chains = 1))
tmp.data <- data.list
tmp.data$y[, 1, 1] <- NA
out <- spMsPGOcc(occ.formula = occ.formula,
det.formula = det.formula,
data = tmp.data,
n.batch = 40,
batch.length = batch.length,
cov.model = "exponential",
tuning = tuning.list,
NNGP = TRUE,
verbose = FALSE,
n.neighbors = 5,
search.type = 'cb',
n.report = 10,
n.burn = 100,
n.chains = 1)
expect_s3_class(out, "spMsPGOcc")
})
# Check default values ----------------
test_that("default priors, inits, burn, thin work", {
out <- spMsPGOcc(occ.formula = occ.formula,
det.formula = det.formula,
data = data.list,
n.batch = n.batch,
batch.length = batch.length,
accept.rate = 0.43,
cov.model = "exponential",
tuning = tuning.list,
NNGP = TRUE,
verbose = FALSE,
n.neighbors = 5,
n.chains = 1)
expect_s3_class(out, "spMsPGOcc")
})
test_that("all correlation functions work", {
out <- spMsPGOcc(occ.formula = occ.formula,
det.formula = det.formula,
data = data.list,
inits = inits.list,
n.batch = n.batch,
batch.length = batch.length,
accept.rate = 0.43,
priors = prior.list,
cov.model = "gaussian",
tuning = list(phi = 0.3),
n.omp.threads = 1,
verbose = FALSE,
NNGP = TRUE,
n.neighbors = 5,
search.type = 'cb',
n.report = 10,
n.burn = 100,
n.thin = 1,
n.chains = 1)
expect_s3_class(out, "spMsPGOcc")
out <- spMsPGOcc(occ.formula = occ.formula,
det.formula = det.formula,
data = data.list,
inits = inits.list,
n.batch = n.batch,
batch.length = batch.length,
accept.rate = 0.43,
priors = prior.list,
cov.model = "spherical",
tuning = list(phi = 0.3),
n.omp.threads = 1,
verbose = FALSE,
NNGP = TRUE,
n.neighbors = 5,
search.type = 'cb',
n.report = 10,
n.burn = 100,
n.thin = 1,
n.chains = 1)
expect_s3_class(out, "spMsPGOcc")
out <- spMsPGOcc(occ.formula = occ.formula,
det.formula = det.formula,
data = data.list,
inits = inits.list,
n.batch = n.batch,
batch.length = batch.length,
accept.rate = 0.43,
priors = list(nu.unif = list(0.4, 3)),
cov.model = "matern",
tuning = list(phi = 0.3, nu = 0.2),
n.omp.threads = 1,
verbose = FALSE,
NNGP = TRUE,
n.neighbors = 5,
search.type = 'cb',
n.report = 10,
n.burn = 100,
n.thin = 1,
n.chains = 1)
expect_s3_class(out, "spMsPGOcc")
})
test_that("verbose prints to the screen", {
expect_output(spMsPGOcc(occ.formula = occ.formula,
det.formula = det.formula,
data = data.list,
inits = inits.list,
n.batch = n.batch,
batch.length = batch.length,
accept.rate = 0.43,
priors = prior.list,
cov.model = "exponential",
tuning = tuning.list,
n.omp.threads = 1,
verbose = TRUE,
NNGP = TRUE,
n.neighbors = 5,
search.type = 'cb',
n.report = 10,
n.burn = 100,
n.thin = 1,
n.chains = 1))
})
# Check waicOcc -----------------------
test_that("waicOCC works for spMsPGOcc", {
# as.vector gets rid of names
waic.out <- as.vector(waicOcc(out))
expect_equal(length(waic.out), 3)
expect_equal(waic.out[3], -2 * (waic.out[1] - waic.out[2]))
})
test_that("fitted works for spMsPGOcc", {
fitted.out <- fitted(out)
expect_equal(length(fitted.out), 2)
expect_equal(class(fitted.out$y.rep.samples), "array")
expect_equal(class(fitted.out$p.samples), "array")
expect_equal(dim(fitted.out$y.rep.samples), dim(fitted.out$p.samples))
})
test_that("predict works for spMsPGOcc", {
n.post.samples <- out$n.post * out$n.chains
X.0 <- cbind(X.0)
colnames(X.0) <- c('int')
pred.out <- predict(out, X.0, coords.0, verbose = FALSE)
expect_type(pred.out, "list")
expect_equal(dim(pred.out$psi.0.samples), c(n.post.samples, N, nrow(X.0)))
expect_equal(dim(pred.out$z.0.samples), c(n.post.samples, N, nrow(X.0)))
})
test_that("detection prediction works", {
X.p.0 <- cbind(dat$X.p[, 1, 1], dat$X.p.re[, 1, 1])
colnames(X.p.0) <- c('intercept', 'det.factor.1')
pred.out <- predict(out, X.p.0, type = 'detection')
expect_type(pred.out, 'list')
expect_equal(dim(pred.out$p.0.samples), c(out$n.post * out$n.chains, N, J))
})
test_that("posterior predictive checks work for msPGOcc", {
n.post.samples <- out$n.post * out$n.chains
J.fit <- nrow(X)
ppc.out <- ppcOcc(out, 'chi-square', 2)
expect_type(ppc.out, "list")
expect_equal(dim(ppc.out$fit.y), c(n.post.samples, N))
expect_equal(dim(ppc.out$fit.y.rep), c(n.post.samples, N))
expect_equal(dim(ppc.out$fit.y.group.quants), c(5, N, max(n.rep)))
expect_equal(dim(ppc.out$fit.y.rep.group.quants), c(5, N, max(n.rep)))
ppc.out <- ppcOcc(out, 'chi-square', 1)
expect_type(ppc.out, "list")
expect_equal(dim(ppc.out$fit.y), c(n.post.samples, N))
expect_equal(dim(ppc.out$fit.y.rep), c(n.post.samples, N))
expect_equal(dim(ppc.out$fit.y.group.quants), c(5, N, J.fit))
expect_equal(dim(ppc.out$fit.y.rep.group.quants), c(5, N, J.fit))
ppc.out <- ppcOcc(out, 'freeman-tukey', 1)
expect_type(ppc.out, "list")
expect_equal(dim(ppc.out$fit.y), c(n.post.samples, N))
expect_equal(dim(ppc.out$fit.y.rep), c(n.post.samples, N))
expect_equal(dim(ppc.out$fit.y.group.quants), c(5, N, J.fit))
expect_equal(dim(ppc.out$fit.y.rep.group.quants), c(5, N, J.fit))
ppc.out <- ppcOcc(out, 'freeman-tukey', 2)
expect_type(ppc.out, "list")
expect_equal(dim(ppc.out$fit.y), c(n.post.samples, N))
expect_equal(dim(ppc.out$fit.y.rep), c(n.post.samples, N))
expect_equal(dim(ppc.out$fit.y.group.quants), c(5, N, max(n.rep)))
expect_equal(dim(ppc.out$fit.y.rep.group.quants), c(5, N, max(n.rep)))
})
# Multiple random intercepts on detection ---------------------------------
J.x <- 8
J.y <- 8
J <- J.x * J.y
n.rep<- sample(2:4, size = J, replace = TRUE)
N <- 6
# Community-level covariate effects
# Occurrence
beta.mean <- c(0.2)
p.occ <- length(beta.mean)
tau.sq.beta <- c(0.6)
# Detection
alpha.mean <- c(0)
tau.sq.alpha <- c(1)
p.det <- length(alpha.mean)
# Random effects
psi.RE <- list()
p.RE <- list(levels = c(50, 10),
sigma.sq.p = c(2.50, 1.5))
# Draw species-level effects from community means.
beta <- matrix(NA, nrow = N, ncol = p.occ)
alpha <- matrix(NA, nrow = N, ncol = p.det)
for (i in 1:p.occ) {
beta[, i] <- rnorm(N, beta.mean[i], sqrt(tau.sq.beta[i]))
}
for (i in 1:p.det) {
alpha[, i] <- rnorm(N, alpha.mean[i], sqrt(tau.sq.alpha[i]))
}
alpha.true <- alpha
phi <- rep(3 / .7, N)
sigma.sq <- rep(2, N)
nu <- rep(2, N)
dat <- simMsOcc(J.x = J.x, J.y = J.y, n.rep = n.rep, N = N, beta = beta, alpha = alpha,
psi.RE = psi.RE, p.RE = p.RE, sp = TRUE, sigma.sq = sigma.sq,
phi = phi, nu = nu, cov.model = 'matern')
pred.indx <- sample(1:J, round(J * .25), replace = FALSE)
y <- dat$y[, -pred.indx, , drop = FALSE]
# Occupancy covariates
X <- dat$X[-pred.indx, , drop = FALSE]
# X.re <- dat$X.re[-pred.indx, , drop = FALSE]
coords <- as.matrix(dat$coords[-pred.indx, , drop = FALSE])
# Prediction covariates
X.0 <- dat$X[pred.indx, , drop = FALSE]
# X.re.0 <- dat$X.re[pred.indx, , drop = FALSE]
coords.0 <- as.matrix(dat$coords[pred.indx, , drop = FALSE])
# Detection covariates
X.p <- dat$X.p[-pred.indx, , , drop = FALSE]
X.p.re <- dat$X.p.re[-pred.indx, , , drop = FALSE]
occ.covs <- cbind(X)
colnames(occ.covs) <- c('int')
det.covs <- list(det.factor.1 = X.p.re[, , 1],
det.factor.2 = X.p.re[, , 2])
data.list <- list(y = y, coords = coords, occ.covs = occ.covs, det.covs = det.covs)
# Priors
prior.list <- list(beta.comm.normal = list(mean = 0, var = 2.72),
alpha.comm.normal = list(mean = 0, var = 2.72),
tau.sq.beta.ig = list(a = 0.1, b = 0.1),
tau.sq.alpha.ig = list(a = 0.1, b = 0.1),
nu.unif = list(0.5, 2.5))
# Starting values
inits.list <- list(alpha.comm = 0,
beta.comm = 0,
beta = 0,
alpha = 0,
tau.sq.beta = 1,
tau.sq.alpha = 1,
z = apply(y, c(1, 2), max, na.rm = TRUE))
# Tuning
tuning.list <- list(phi = 1, nu = 0.25)
batch.length <- 25
n.batch <- 10
n.report <- 100
occ.formula <- ~ 1
det.formula <- ~ (1 | det.factor.1) + (1 | det.factor.2)
out <- spMsPGOcc(occ.formula = occ.formula,
det.formula = det.formula,
data = data.list,
inits = inits.list,
n.batch = n.batch,
batch.length = batch.length,
accept.rate = 0.43,
priors = prior.list,
cov.model = "matern",
tuning = tuning.list,
n.omp.threads = 1,
verbose = FALSE,
NNGP = TRUE,
n.neighbors = 5,
search.type = 'cb',
n.report = 10,
n.burn = 100,
n.thin = 2,
n.chains = 2,
k.fold = 2,
k.fold.threads = 1)
# To make sure it worked --------------
test_that("out is of class spMsPGOcc", {
expect_s3_class(out, "spMsPGOcc")
})
# Check cross-validation --------------
test_that("cross-validation works", {
expect_equal(length(out$k.fold.deviance), N)
expect_type(out$k.fold.deviance, "double")
expect_equal(sum(out$k.fold.deviance < 0), 0)
})
# Check random effects ----------------
test_that("random effects are correct", {
expect_equal(out$pRE, TRUE)
expect_equal(out$psiRE, FALSE)
})
# Check RE error ----------------------
test_that("random effect gives error when non-numeric", {
# data.list$occ.covs <- as.data.frame(data.list$occ.covs)
# data.list$occ.covs$occ.factor.1 <- factor(data.list$occ.covs$occ.factor.1)
data.list$det.covs$det.factor.1 <- factor(data.list$det.covs$det.factor.1)
expect_error(out <- spMsPGOcc(occ.formula = occ.formula,
det.formula = det.formula,
data = data.list,
inits = inits.list,
n.batch = n.batch,
batch.length = batch.length,
accept.rate = 0.43,
priors = prior.list,
cov.model = "matern",
tuning = tuning.list,
n.omp.threads = 1,
verbose = FALSE,
NNGP = TRUE,
n.neighbors = 5,
search.type = 'cb',
n.report = 10,
n.burn = 100,
n.thin = 2,
n.chains = 1))
# data.list$occ.covs$occ.factor.1 <- as.character(factor(data.list$occ.covs$occ.factor.1))
data.list$det.covs$det.factor.1 <- as.character(factor(data.list$det.covs$det.factor.1))
expect_error(out <- spMsPGOcc(occ.formula = occ.formula,
det.formula = det.formula,
data = data.list,
inits = inits.list,
n.batch = n.batch,
batch.length = batch.length,
accept.rate = 0.43,
priors = prior.list,
cov.model = "matern",
tuning = tuning.list,
n.omp.threads = 1,
verbose = FALSE,
NNGP = TRUE,
n.neighbors = 5,
search.type = 'cb',
n.report = 10,
n.burn = 100,
n.thin = 2,
n.chains = 1))
})
# Check RE levels ---------------------
test_that("random effect levels are correct", {
expect_equal(sort(unique(unlist(out$p.re.level.names))),
sort(unique(c(X.p.re))))
})
# Check output data output is correct -
test_that("out$y == y", {
expect_equal(out$y, y)
})
# Check missing values ----------------
test_that("missing value error handling works", {
# tmp.data <- data.list
# tmp.data$occ.covs[3, ] <- NA
# expect_error(spMsPGOcc(occ.formula = occ.formula,
# det.formula = det.formula,
# data = tmp.data,
# n.batch = 40,
# batch.length = batch.length,
# cov.model = "exponential",
# tuning = tuning.list,
# NNGP = TRUE,
# verbose = FALSE,
# n.neighbors = 5,
# search.type = 'cb',
# n.report = 10,
# n.burn = 100,
# n.chains = 1))
tmp.data <- data.list
tmp.data$y[1, 1, 1] <- 1
tmp.data$det.covs[[1]][1] <- NA
expect_error(spMsPGOcc(occ.formula = occ.formula,
det.formula = det.formula,
data = tmp.data,
n.batch = 40,
batch.length = batch.length,
cov.model = "exponential",
tuning = tuning.list,
NNGP = TRUE,
verbose = FALSE,
n.neighbors = 5,
search.type = 'cb',
n.report = 10,
n.burn = 100,
n.chains = 1))
tmp.data <- data.list
tmp.data$y[, 1, 1] <- NA
out <- spMsPGOcc(occ.formula = occ.formula,
det.formula = det.formula,
data = tmp.data,
n.batch = 40,
batch.length = batch.length,
cov.model = "exponential",
tuning = tuning.list,
NNGP = TRUE,
verbose = FALSE,
n.neighbors = 5,
search.type = 'cb',
n.report = 10,
n.burn = 100,
n.chains = 1)
expect_s3_class(out, "spMsPGOcc")
})
# Check default values ----------------
test_that("default priors, inits, burn, thin work", {
out <- spMsPGOcc(occ.formula = occ.formula,
det.formula = det.formula,
data = data.list,
n.batch = n.batch,
batch.length = batch.length,
accept.rate = 0.43,
cov.model = "exponential",
tuning = tuning.list,
NNGP = TRUE,
verbose = FALSE,
n.neighbors = 5,
n.chains = 1)
expect_s3_class(out, "spMsPGOcc")
})
test_that("all correlation functions work", {
out <- spMsPGOcc(occ.formula = occ.formula,
det.formula = det.formula,
data = data.list,
inits = inits.list,
n.batch = n.batch,
batch.length = batch.length,
accept.rate = 0.43,
priors = prior.list,
cov.model = "gaussian",
tuning = list(phi = 0.3),
n.omp.threads = 1,
verbose = FALSE,
NNGP = TRUE,
n.neighbors = 5,
search.type = 'cb',
n.report = 10,
n.burn = 100,
n.thin = 1,
n.chains = 1)
expect_s3_class(out, "spMsPGOcc")
out <- spMsPGOcc(occ.formula = occ.formula,
det.formula = det.formula,
data = data.list,
inits = inits.list,
n.batch = n.batch,
batch.length = batch.length,
accept.rate = 0.43,
priors = prior.list,
cov.model = "spherical",
tuning = list(phi = 0.3),
n.omp.threads = 1,
verbose = FALSE,
NNGP = TRUE,
n.neighbors = 5,
search.type = 'cb',
n.report = 10,
n.burn = 100,
n.thin = 1,
n.chains = 1)
expect_s3_class(out, "spMsPGOcc")
out <- spMsPGOcc(occ.formula = occ.formula,
det.formula = det.formula,
data = data.list,
inits = inits.list,
n.batch = n.batch,
batch.length = batch.length,
accept.rate = 0.43,
priors = list(nu.unif = list(0.4, 3)),
cov.model = "matern",
tuning = list(phi = 0.3, nu = 0.2),
n.omp.threads = 1,
verbose = FALSE,
NNGP = TRUE,
n.neighbors = 5,
search.type = 'cb',
n.report = 10,
n.burn = 100,
n.thin = 1,
n.chains = 1)
expect_s3_class(out, "spMsPGOcc")
})
test_that("verbose prints to the screen", {
expect_output(spMsPGOcc(occ.formula = occ.formula,
det.formula = det.formula,
data = data.list,
inits = inits.list,
n.batch = n.batch,
batch.length = batch.length,
accept.rate = 0.43,
priors = prior.list,
cov.model = "exponential",
tuning = tuning.list,
n.omp.threads = 1,
verbose = TRUE,
NNGP = TRUE,
n.neighbors = 5,
search.type = 'cb',
n.report = 10,
n.burn = 100,
n.thin = 1,
n.chains = 1))
})
# Check waicOcc -----------------------
test_that("waicOCC works for spMsPGOcc", {
# as.vector gets rid of names
waic.out <- as.vector(waicOcc(out))
expect_equal(length(waic.out), 3)
expect_equal(waic.out[3], -2 * (waic.out[1] - waic.out[2]))
})
test_that("fitted works for spMsPGOcc", {
fitted.out <- fitted(out)
expect_equal(length(fitted.out), 2)
expect_equal(class(fitted.out$y.rep.samples), "array")
expect_equal(class(fitted.out$p.samples), "array")
expect_equal(dim(fitted.out$y.rep.samples), dim(fitted.out$p.samples))
})
test_that("predict works for spMsPGOcc", {
n.post.samples <- out$n.post * out$n.chains
X.0 <- cbind(X.0)
colnames(X.0) <- c('int')
pred.out <- predict(out, X.0, coords.0, verbose = FALSE)
expect_type(pred.out, "list")
expect_equal(dim(pred.out$psi.0.samples), c(n.post.samples, N, nrow(X.0)))
expect_equal(dim(pred.out$z.0.samples), c(n.post.samples, N, nrow(X.0)))
})
test_that("detection prediction works", {
X.p.0 <- cbind(dat$X.p[, 1, 1], dat$X.p.re[, 1, 1:2])
colnames(X.p.0) <- c('intercept', 'det.factor.1', 'det.factor.2')
pred.out <- predict(out, X.p.0, type = 'detection')
expect_type(pred.out, 'list')
expect_equal(dim(pred.out$p.0.samples), c(out$n.post * out$n.chains, N, J))
})
test_that("posterior predictive checks work for msPGOcc", {
n.post.samples <- out$n.post * out$n.chains
J.fit <- nrow(X)
ppc.out <- ppcOcc(out, 'chi-square', 2)
expect_type(ppc.out, "list")
expect_equal(dim(ppc.out$fit.y), c(n.post.samples, N))
expect_equal(dim(ppc.out$fit.y.rep), c(n.post.samples, N))
expect_equal(dim(ppc.out$fit.y.group.quants), c(5, N, max(n.rep)))
expect_equal(dim(ppc.out$fit.y.rep.group.quants), c(5, N, max(n.rep)))
ppc.out <- ppcOcc(out, 'chi-square', 1)
expect_type(ppc.out, "list")
expect_equal(dim(ppc.out$fit.y), c(n.post.samples, N))
expect_equal(dim(ppc.out$fit.y.rep), c(n.post.samples, N))
expect_equal(dim(ppc.out$fit.y.group.quants), c(5, N, J.fit))
expect_equal(dim(ppc.out$fit.y.rep.group.quants), c(5, N, J.fit))
ppc.out <- ppcOcc(out, 'freeman-tukey', 1)
expect_type(ppc.out, "list")
expect_equal(dim(ppc.out$fit.y), c(n.post.samples, N))
expect_equal(dim(ppc.out$fit.y.rep), c(n.post.samples, N))
expect_equal(dim(ppc.out$fit.y.group.quants), c(5, N, J.fit))
expect_equal(dim(ppc.out$fit.y.rep.group.quants), c(5, N, J.fit))
ppc.out <- ppcOcc(out, 'freeman-tukey', 2)
expect_type(ppc.out, "list")
expect_equal(dim(ppc.out$fit.y), c(n.post.samples, N))
expect_equal(dim(ppc.out$fit.y.rep), c(n.post.samples, N))
expect_equal(dim(ppc.out$fit.y.group.quants), c(5, N, max(n.rep)))
expect_equal(dim(ppc.out$fit.y.rep.group.quants), c(5, N, max(n.rep)))
})
# Detection random effects with covariates --------------------------------
J.x <- 8
J.y <- 8
J <- J.x * J.y
n.rep<- sample(2:4, size = J, replace = TRUE)
N <- 6
# Community-level covariate effects
# Occurrence
beta.mean <- c(0.2)
p.occ <- length(beta.mean)
tau.sq.beta <- c(0.6)
# Detection
alpha.mean <- c(0, 0.5)
tau.sq.alpha <- c(1, 2.3)
p.det <- length(alpha.mean)
# Random effects
psi.RE <- list()
p.RE <- list(levels = c(50, 10),
sigma.sq.p = c(2.50, 1.5))
# Draw species-level effects from community means.
beta <- matrix(NA, nrow = N, ncol = p.occ)
alpha <- matrix(NA, nrow = N, ncol = p.det)
for (i in 1:p.occ) {
beta[, i] <- rnorm(N, beta.mean[i], sqrt(tau.sq.beta[i]))
}
for (i in 1:p.det) {
alpha[, i] <- rnorm(N, alpha.mean[i], sqrt(tau.sq.alpha[i]))
}
alpha.true <- alpha
phi <- rep(3 / .7, N)
sigma.sq <- rep(2, N)
nu <- rep(2, N)
dat <- simMsOcc(J.x = J.x, J.y = J.y, n.rep = n.rep, N = N, beta = beta, alpha = alpha,
psi.RE = psi.RE, p.RE = p.RE, sp = TRUE, sigma.sq = sigma.sq,
phi = phi, nu = nu, cov.model = 'matern')
pred.indx <- sample(1:J, round(J * .25), replace = FALSE)
y <- dat$y[, -pred.indx, , drop = FALSE]
# Occupancy covariates
X <- dat$X[-pred.indx, , drop = FALSE]
# X.re <- dat$X.re[-pred.indx, , drop = FALSE]
coords <- as.matrix(dat$coords[-pred.indx, , drop = FALSE])
# Prediction covariates
X.0 <- dat$X[pred.indx, , drop = FALSE]
# X.re.0 <- dat$X.re[pred.indx, , drop = FALSE]
coords.0 <- as.matrix(dat$coords[pred.indx, , drop = FALSE])
# Detection covariates
X.p <- dat$X.p[-pred.indx, , , drop = FALSE]
X.p.re <- dat$X.p.re[-pred.indx, , , drop = FALSE]
occ.covs <- cbind(X)
colnames(occ.covs) <- c('int')
det.covs <- list(det.cov.1 = X.p[, , 2],
det.factor.1 = X.p.re[, , 1],
det.factor.2 = X.p.re[, , 2])
data.list <- list(y = y, coords = coords, occ.covs = occ.covs, det.covs = det.covs)
# Priors
prior.list <- list(beta.comm.normal = list(mean = 0, var = 2.72),
alpha.comm.normal = list(mean = 0, var = 2.72),
tau.sq.beta.ig = list(a = 0.1, b = 0.1),
tau.sq.alpha.ig = list(a = 0.1, b = 0.1),
nu.unif = list(0.5, 2.5))
# Starting values
inits.list <- list(alpha.comm = 0,
beta.comm = 0,
beta = 0,
alpha = 0,
tau.sq.beta = 1,
tau.sq.alpha = 1,
z = apply(y, c(1, 2), max, na.rm = TRUE))
# Tuning
tuning.list <- list(phi = 1, nu = 0.25)
batch.length <- 25
n.batch <- 10
n.report <- 100
occ.formula <- ~ 1
det.formula <- ~ det.cov.1 + (1 | det.factor.1) + (1 | det.factor.2)
out <- spMsPGOcc(occ.formula = occ.formula,
det.formula = det.formula,
data = data.list,
inits = inits.list,
n.batch = n.batch,
batch.length = batch.length,
accept.rate = 0.43,
priors = prior.list,
cov.model = "matern",
tuning = tuning.list,
n.omp.threads = 1,
verbose = FALSE,
NNGP = TRUE,
n.neighbors = 5,
search.type = 'cb',
n.report = 10,
n.burn = 100,
n.thin = 2,
n.chains = 2,
k.fold = 2,
k.fold.threads = 1)
# To make sure it worked --------------
test_that("out is of class spMsPGOcc", {
expect_s3_class(out, "spMsPGOcc")
})
# Check cross-validation --------------
test_that("cross-validation works", {
expect_equal(length(out$k.fold.deviance), N)
expect_type(out$k.fold.deviance, "double")
expect_equal(sum(out$k.fold.deviance < 0), 0)
})
# Check random effects ----------------
test_that("random effects are correct", {
expect_equal(out$pRE, TRUE)
expect_equal(out$psiRE, FALSE)
})
# Check RE error ----------------------
test_that("random effect gives error when non-numeric", {
# data.list$occ.covs <- as.data.frame(data.list$occ.covs)
# data.list$occ.covs$occ.factor.1 <- factor(data.list$occ.covs$occ.factor.1)
data.list$det.covs$det.factor.1 <- factor(data.list$det.covs$det.factor.1)
expect_error(out <- spMsPGOcc(occ.formula = occ.formula,
det.formula = det.formula,
data = data.list,
inits = inits.list,
n.batch = n.batch,
batch.length = batch.length,
accept.rate = 0.43,
priors = prior.list,
cov.model = "matern",
tuning = tuning.list,
n.omp.threads = 1,
verbose = FALSE,
NNGP = TRUE,
n.neighbors = 5,
search.type = 'cb',
n.report = 10,
n.burn = 100,
n.thin = 2,
n.chains = 1))
# data.list$occ.covs$occ.factor.1 <- as.character(factor(data.list$occ.covs$occ.factor.1))
data.list$det.covs$det.factor.1 <- as.character(factor(data.list$det.covs$det.factor.1))
expect_error(out <- spMsPGOcc(occ.formula = occ.formula,
det.formula = det.formula,
data = data.list,
inits = inits.list,
n.batch = n.batch,
batch.length = batch.length,
accept.rate = 0.43,
priors = prior.list,
cov.model = "matern",
tuning = tuning.list,
n.omp.threads = 1,
verbose = FALSE,
NNGP = TRUE,
n.neighbors = 5,
search.type = 'cb',
n.report = 10,
n.burn = 100,
n.thin = 2,
n.chains = 1))
})
# Check RE levels ---------------------
test_that("random effect levels are correct", {
expect_equal(sort(unique(unlist(out$p.re.level.names))),
sort(unique(c(X.p.re))))
})
# Check output data output is correct -
test_that("out$y == y", {
expect_equal(out$y, y)
})
# Check missing values ----------------
test_that("missing value error handling works", {
# tmp.data <- data.list
# tmp.data$occ.covs[3, ] <- NA
# expect_error(spMsPGOcc(occ.formula = occ.formula,
# det.formula = det.formula,
# data = tmp.data,
# n.batch = 40,
# batch.length = batch.length,
# cov.model = "exponential",
# tuning = tuning.list,
# NNGP = TRUE,
# verbose = FALSE,
# n.neighbors = 5,
# search.type = 'cb',
# n.report = 10,
# n.burn = 100,
# n.chains = 1))
tmp.data <- data.list
tmp.data$y[1, 1, 1] <- 1
tmp.data$det.covs[[1]][1] <- NA
expect_error(spMsPGOcc(occ.formula = occ.formula,
det.formula = det.formula,
data = tmp.data,
n.batch = 40,
batch.length = batch.length,
cov.model = "exponential",
tuning = tuning.list,
NNGP = TRUE,
verbose = FALSE,
n.neighbors = 5,
search.type = 'cb',
n.report = 10,
n.burn = 100,
n.chains = 1))
tmp.data <- data.list
tmp.data$y[, 1, 1] <- NA
out <- spMsPGOcc(occ.formula = occ.formula,
det.formula = det.formula,
data = tmp.data,
n.batch = 40,
batch.length = batch.length,
cov.model = "exponential",
tuning = tuning.list,
NNGP = TRUE,
verbose = FALSE,
n.neighbors = 5,
search.type = 'cb',
n.report = 10,
n.burn = 100,
n.chains = 1)
expect_s3_class(out, "spMsPGOcc")
})
# Check default values ----------------
test_that("default priors, inits, burn, thin work", {
out <- spMsPGOcc(occ.formula = occ.formula,
det.formula = det.formula,
data = data.list,
n.batch = n.batch,
batch.length = batch.length,
accept.rate = 0.43,
cov.model = "exponential",
tuning = tuning.list,
NNGP = TRUE,
verbose = FALSE,
n.neighbors = 5,
n.chains = 1)
expect_s3_class(out, "spMsPGOcc")
})
test_that("all correlation functions work", {
out <- spMsPGOcc(occ.formula = occ.formula,
det.formula = det.formula,
data = data.list,
inits = inits.list,
n.batch = n.batch,
batch.length = batch.length,
accept.rate = 0.43,
priors = prior.list,
cov.model = "gaussian",
tuning = list(phi = 0.3),
n.omp.threads = 1,
verbose = FALSE,
NNGP = TRUE,
n.neighbors = 5,
search.type = 'cb',
n.report = 10,
n.burn = 100,
n.thin = 1,
n.chains = 1)
expect_s3_class(out, "spMsPGOcc")
out <- spMsPGOcc(occ.formula = occ.formula,
det.formula = det.formula,
data = data.list,
inits = inits.list,
n.batch = n.batch,
batch.length = batch.length,
accept.rate = 0.43,
priors = prior.list,
cov.model = "spherical",
tuning = list(phi = 0.3),
n.omp.threads = 1,
verbose = FALSE,
NNGP = TRUE,
n.neighbors = 5,
search.type = 'cb',
n.report = 10,
n.burn = 100,
n.thin = 1,
n.chains = 1)
expect_s3_class(out, "spMsPGOcc")
out <- spMsPGOcc(occ.formula = occ.formula,
det.formula = det.formula,
data = data.list,
inits = inits.list,
n.batch = n.batch,
batch.length = batch.length,
accept.rate = 0.43,
priors = list(nu.unif = list(0.4, 3)),
cov.model = "matern",
tuning = list(phi = 0.3, nu = 0.2),
n.omp.threads = 1,
verbose = FALSE,
NNGP = TRUE,
n.neighbors = 5,
search.type = 'cb',
n.report = 10,
n.burn = 100,
n.thin = 1,
n.chains = 1)
expect_s3_class(out, "spMsPGOcc")
})
test_that("verbose prints to the screen", {
expect_output(spMsPGOcc(occ.formula = occ.formula,
det.formula = det.formula,
data = data.list,
inits = inits.list,
n.batch = n.batch,
batch.length = batch.length,
accept.rate = 0.43,
priors = prior.list,
cov.model = "exponential",
tuning = tuning.list,
n.omp.threads = 1,
verbose = TRUE,
NNGP = TRUE,
n.neighbors = 5,
search.type = 'cb',
n.report = 10,
n.burn = 100,
n.thin = 1,
n.chains = 1))
})
# Check waicOcc -----------------------
test_that("waicOCC works for spMsPGOcc", {
# as.vector gets rid of names
waic.out <- as.vector(waicOcc(out))
expect_equal(length(waic.out), 3)
expect_equal(waic.out[3], -2 * (waic.out[1] - waic.out[2]))
})
test_that("fitted works for spMsPGOcc", {
fitted.out <- fitted(out)
expect_equal(length(fitted.out), 2)
expect_equal(class(fitted.out$y.rep.samples), "array")
expect_equal(class(fitted.out$p.samples), "array")
expect_equal(dim(fitted.out$y.rep.samples), dim(fitted.out$p.samples))
})
test_that("predict works for spMsPGOcc", {
n.post.samples <- out$n.post * out$n.chains
X.0 <- cbind(X.0)
colnames(X.0) <- c('int')
pred.out <- predict(out, X.0, coords.0, verbose = FALSE)
expect_type(pred.out, "list")
expect_equal(dim(pred.out$psi.0.samples), c(n.post.samples, N, nrow(X.0)))
expect_equal(dim(pred.out$z.0.samples), c(n.post.samples, N, nrow(X.0)))
})
test_that("detection prediction works", {
X.p.0 <- cbind(dat$X.p[, 1, ], dat$X.p.re[, 1, 1:2])
colnames(X.p.0) <- c('intercept', 'det.cov.1', 'det.factor.1', 'det.factor.2')
pred.out <- predict(out, X.p.0, type = 'detection')
expect_type(pred.out, 'list')
expect_equal(dim(pred.out$p.0.samples), c(out$n.post * out$n.chains, N, J))
})
test_that("posterior predictive checks work for msPGOcc", {
n.post.samples <- out$n.post * out$n.chains
J.fit <- nrow(X)
ppc.out <- ppcOcc(out, 'chi-square', 2)
expect_type(ppc.out, "list")
expect_equal(dim(ppc.out$fit.y), c(n.post.samples, N))
expect_equal(dim(ppc.out$fit.y.rep), c(n.post.samples, N))
expect_equal(dim(ppc.out$fit.y.group.quants), c(5, N, max(n.rep)))
expect_equal(dim(ppc.out$fit.y.rep.group.quants), c(5, N, max(n.rep)))
ppc.out <- ppcOcc(out, 'chi-square', 1)
expect_type(ppc.out, "list")
expect_equal(dim(ppc.out$fit.y), c(n.post.samples, N))
expect_equal(dim(ppc.out$fit.y.rep), c(n.post.samples, N))
expect_equal(dim(ppc.out$fit.y.group.quants), c(5, N, J.fit))
expect_equal(dim(ppc.out$fit.y.rep.group.quants), c(5, N, J.fit))
ppc.out <- ppcOcc(out, 'freeman-tukey', 1)
expect_type(ppc.out, "list")
expect_equal(dim(ppc.out$fit.y), c(n.post.samples, N))
expect_equal(dim(ppc.out$fit.y.rep), c(n.post.samples, N))
expect_equal(dim(ppc.out$fit.y.group.quants), c(5, N, J.fit))
expect_equal(dim(ppc.out$fit.y.rep.group.quants), c(5, N, J.fit))
ppc.out <- ppcOcc(out, 'freeman-tukey', 2)
expect_type(ppc.out, "list")
expect_equal(dim(ppc.out$fit.y), c(n.post.samples, N))
expect_equal(dim(ppc.out$fit.y.rep), c(n.post.samples, N))
expect_equal(dim(ppc.out$fit.y.group.quants), c(5, N, max(n.rep)))
expect_equal(dim(ppc.out$fit.y.rep.group.quants), c(5, N, max(n.rep)))
})
# Detection random effects with covariates on all -------------------------
J.x <- 8
J.y <- 8
J <- J.x * J.y
n.rep<- sample(2:4, size = J, replace = TRUE)
N <- 6
# Community-level covariate effects
# Occurrence
beta.mean <- c(0.2, 1.5, 0.3)
p.occ <- length(beta.mean)
tau.sq.beta <- c(0.6, 0.5, 3.3)
# Detection
alpha.mean <- c(0, 0.5)
tau.sq.alpha <- c(1, 2.3)
p.det <- length(alpha.mean)
# Random effects
psi.RE <- list()
p.RE <- list(levels = c(50, 10),
sigma.sq.p = c(2.50, 1.5))
# Draw species-level effects from community means.
beta <- matrix(NA, nrow = N, ncol = p.occ)
alpha <- matrix(NA, nrow = N, ncol = p.det)
for (i in 1:p.occ) {
beta[, i] <- rnorm(N, beta.mean[i], sqrt(tau.sq.beta[i]))
}
for (i in 1:p.det) {
alpha[, i] <- rnorm(N, alpha.mean[i], sqrt(tau.sq.alpha[i]))
}
alpha.true <- alpha
phi <- rep(3 / .7, N)
sigma.sq <- rep(2, N)
nu <- rep(2, N)
dat <- simMsOcc(J.x = J.x, J.y = J.y, n.rep = n.rep, N = N, beta = beta, alpha = alpha,
psi.RE = psi.RE, p.RE = p.RE, sp = TRUE, sigma.sq = sigma.sq,
phi = phi, nu = nu, cov.model = 'matern')
pred.indx <- sample(1:J, round(J * .25), replace = FALSE)
y <- dat$y[, -pred.indx, , drop = FALSE]
# Occupancy covariates
X <- dat$X[-pred.indx, , drop = FALSE]
# X.re <- dat$X.re[-pred.indx, , drop = FALSE]
coords <- as.matrix(dat$coords[-pred.indx, , drop = FALSE])
# Prediction covariates
X.0 <- dat$X[pred.indx, , drop = FALSE]
# X.re.0 <- dat$X.re[pred.indx, , drop = FALSE]
coords.0 <- as.matrix(dat$coords[pred.indx, , drop = FALSE])
# Detection covariates
X.p <- dat$X.p[-pred.indx, , , drop = FALSE]
X.p.re <- dat$X.p.re[-pred.indx, , , drop = FALSE]
occ.covs <- cbind(X)
colnames(occ.covs) <- c('int', 'occ.cov.1', 'occ.cov.2')
det.covs <- list(det.cov.1 = X.p[, , 2],
det.factor.1 = X.p.re[, , 1],
det.factor.2 = X.p.re[, , 2])
data.list <- list(y = y, coords = coords, occ.covs = occ.covs, det.covs = det.covs)
# Priors
prior.list <- list(beta.comm.normal = list(mean = 0, var = 2.72),
alpha.comm.normal = list(mean = 0, var = 2.72),
tau.sq.beta.ig = list(a = 0.1, b = 0.1),
tau.sq.alpha.ig = list(a = 0.1, b = 0.1),
nu.unif = list(0.5, 2.5))
# Starting values
inits.list <- list(alpha.comm = 0,
beta.comm = 0,
beta = 0,
alpha = 0,
tau.sq.beta = 1,
tau.sq.alpha = 1,
z = apply(y, c(1, 2), max, na.rm = TRUE))
# Tuning
tuning.list <- list(phi = 1, nu = 0.25)
batch.length <- 25
n.batch <- 10
n.report <- 100
occ.formula <- ~ occ.cov.1 + occ.cov.2
det.formula <- ~ det.cov.1 + (1 | det.factor.1) + (1 | det.factor.2)
out <- spMsPGOcc(occ.formula = occ.formula,
det.formula = det.formula,
data = data.list,
inits = inits.list,
n.batch = n.batch,
batch.length = batch.length,
accept.rate = 0.43,
priors = prior.list,
cov.model = "matern",
tuning = tuning.list,
n.omp.threads = 1,
verbose = FALSE,
NNGP = TRUE,
n.neighbors = 5,
search.type = 'cb',
n.report = 10,
n.burn = 100,
n.thin = 2,
n.chains = 2,
k.fold = 2,
k.fold.threads = 1)
# To make sure it worked --------------
test_that("out is of class spMsPGOcc", {
expect_s3_class(out, "spMsPGOcc")
})
# Check cross-validation --------------
test_that("cross-validation works", {
expect_equal(length(out$k.fold.deviance), N)
expect_type(out$k.fold.deviance, "double")
expect_equal(sum(out$k.fold.deviance < 0), 0)
})
# Check random effects ----------------
test_that("random effects are correct", {
expect_equal(out$pRE, TRUE)
expect_equal(out$psiRE, FALSE)
})
# Check RE error ----------------------
test_that("random effect gives error when non-numeric", {
# data.list$occ.covs <- as.data.frame(data.list$occ.covs)
# data.list$occ.covs$occ.factor.1 <- factor(data.list$occ.covs$occ.factor.1)
data.list$det.covs$det.factor.1 <- factor(data.list$det.covs$det.factor.1)
expect_error(out <- spMsPGOcc(occ.formula = occ.formula,
det.formula = det.formula,
data = data.list,
inits = inits.list,
n.batch = n.batch,
batch.length = batch.length,
accept.rate = 0.43,
priors = prior.list,
cov.model = "matern",
tuning = tuning.list,
n.omp.threads = 1,
verbose = FALSE,
NNGP = TRUE,
n.neighbors = 5,
search.type = 'cb',
n.report = 10,
n.burn = 100,
n.thin = 2,
n.chains = 1))
# data.list$occ.covs$occ.factor.1 <- as.character(factor(data.list$occ.covs$occ.factor.1))
data.list$det.covs$det.factor.1 <- as.character(factor(data.list$det.covs$det.factor.1))
expect_error(out <- spMsPGOcc(occ.formula = occ.formula,
det.formula = det.formula,
data = data.list,
inits = inits.list,
n.batch = n.batch,
batch.length = batch.length,
accept.rate = 0.43,
priors = prior.list,
cov.model = "matern",
tuning = tuning.list,
n.omp.threads = 1,
verbose = FALSE,
NNGP = TRUE,
n.neighbors = 5,
search.type = 'cb',
n.report = 10,
n.burn = 100,
n.thin = 2,
n.chains = 1))
})
# Check RE levels ---------------------
test_that("random effect levels are correct", {
expect_equal(sort(unique(unlist(out$p.re.level.names))),
sort(unique(c(X.p.re))))
})
# Check output data output is correct -
test_that("out$y == y", {
expect_equal(out$y, y)
})
# Check missing values ----------------
test_that("missing value error handling works", {
tmp.data <- data.list
tmp.data$occ.covs[3, ] <- NA
expect_error(spMsPGOcc(occ.formula = occ.formula,
det.formula = det.formula,
data = tmp.data,
n.batch = 40,
batch.length = batch.length,
cov.model = "exponential",
tuning = tuning.list,
NNGP = TRUE,
verbose = FALSE,
n.neighbors = 5,
search.type = 'cb',
n.report = 10,
n.burn = 100,
n.chains = 1))
tmp.data <- data.list
tmp.data$y[1, 1, 1] <- 1
tmp.data$det.covs[[1]][1] <- NA
expect_error(spMsPGOcc(occ.formula = occ.formula,
det.formula = det.formula,
data = tmp.data,
n.batch = 40,
batch.length = batch.length,
cov.model = "exponential",
tuning = tuning.list,
NNGP = TRUE,
verbose = FALSE,
n.neighbors = 5,
search.type = 'cb',
n.report = 10,
n.burn = 100,
n.chains = 1))
tmp.data <- data.list
tmp.data$y[, 1, 1] <- NA
out <- spMsPGOcc(occ.formula = occ.formula,
det.formula = det.formula,
data = tmp.data,
n.batch = 40,
batch.length = batch.length,
cov.model = "exponential",
tuning = tuning.list,
NNGP = TRUE,
verbose = FALSE,
n.neighbors = 5,
search.type = 'cb',
n.report = 10,
n.burn = 100,
n.chains = 1)
expect_s3_class(out, "spMsPGOcc")
})
# Check default values ----------------
test_that("default priors, inits, burn, thin work", {
out <- spMsPGOcc(occ.formula = occ.formula,
det.formula = det.formula,
data = data.list,
n.batch = n.batch,
batch.length = batch.length,
accept.rate = 0.43,
cov.model = "exponential",
tuning = tuning.list,
NNGP = TRUE,
verbose = FALSE,
n.neighbors = 5,
n.chains = 1)
expect_s3_class(out, "spMsPGOcc")
})
test_that("all correlation functions work", {
out <- spMsPGOcc(occ.formula = occ.formula,
det.formula = det.formula,
data = data.list,
inits = inits.list,
n.batch = n.batch,
batch.length = batch.length,
accept.rate = 0.43,
priors = prior.list,
cov.model = "gaussian",
tuning = list(phi = 0.3),
n.omp.threads = 1,
verbose = FALSE,
NNGP = TRUE,
n.neighbors = 5,
search.type = 'cb',
n.report = 10,
n.burn = 100,
n.thin = 1,
n.chains = 1)
expect_s3_class(out, "spMsPGOcc")
out <- spMsPGOcc(occ.formula = occ.formula,
det.formula = det.formula,
data = data.list,
inits = inits.list,
n.batch = n.batch,
batch.length = batch.length,
accept.rate = 0.43,
priors = prior.list,
cov.model = "spherical",
tuning = list(phi = 0.3),
n.omp.threads = 1,
verbose = FALSE,
NNGP = TRUE,
n.neighbors = 5,
search.type = 'cb',
n.report = 10,
n.burn = 100,
n.thin = 1,
n.chains = 1)
expect_s3_class(out, "spMsPGOcc")
out <- spMsPGOcc(occ.formula = occ.formula,
det.formula = det.formula,
data = data.list,
inits = inits.list,
n.batch = n.batch,
batch.length = batch.length,
accept.rate = 0.43,
priors = list(nu.unif = list(0.4, 3)),
cov.model = "matern",
tuning = list(phi = 0.3, nu = 0.2),
n.omp.threads = 1,
verbose = FALSE,
NNGP = TRUE,
n.neighbors = 5,
search.type = 'cb',
n.report = 10,
n.burn = 100,
n.thin = 1,
n.chains = 1)
expect_s3_class(out, "spMsPGOcc")
})
test_that("verbose prints to the screen", {
expect_output(spMsPGOcc(occ.formula = occ.formula,
det.formula = det.formula,
data = data.list,
inits = inits.list,
n.batch = n.batch,
batch.length = batch.length,
accept.rate = 0.43,
priors = prior.list,
cov.model = "exponential",
tuning = tuning.list,
n.omp.threads = 1,
verbose = TRUE,
NNGP = TRUE,
n.neighbors = 5,
search.type = 'cb',
n.report = 10,
n.burn = 100,
n.thin = 1,
n.chains = 1))
})
# Check waicOcc -----------------------
test_that("waicOCC works for spMsPGOcc", {
# as.vector gets rid of names
waic.out <- as.vector(waicOcc(out))
expect_equal(length(waic.out), 3)
expect_equal(waic.out[3], -2 * (waic.out[1] - waic.out[2]))
})
test_that("fitted works for spMsPGOcc", {
fitted.out <- fitted(out)
expect_equal(length(fitted.out), 2)
expect_equal(class(fitted.out$y.rep.samples), "array")
expect_equal(class(fitted.out$p.samples), "array")
expect_equal(dim(fitted.out$y.rep.samples), dim(fitted.out$p.samples))
})
test_that("predict works for spMsPGOcc", {
n.post.samples <- out$n.post * out$n.chains
X.0 <- cbind(X.0)
colnames(X.0) <- c('int', 'occ.cov.1', 'occ.cov.2')
pred.out <- predict(out, X.0, coords.0, verbose = FALSE)
expect_type(pred.out, "list")
expect_equal(dim(pred.out$psi.0.samples), c(n.post.samples, N, nrow(X.0)))
expect_equal(dim(pred.out$z.0.samples), c(n.post.samples, N, nrow(X.0)))
})
test_that("detection prediction works", {
X.p.0 <- cbind(dat$X.p[, 1, ], dat$X.p.re[, 1, 1:2])
colnames(X.p.0) <- c('intercept', 'det.cov.1', 'det.factor.1', 'det.factor.2')
pred.out <- predict(out, X.p.0, type = 'detection')
expect_type(pred.out, 'list')
expect_equal(dim(pred.out$p.0.samples), c(out$n.post * out$n.chains, N, J))
})
test_that("posterior predictive checks work for msPGOcc", {
n.post.samples <- out$n.post * out$n.chains
J.fit <- nrow(X)
ppc.out <- ppcOcc(out, 'chi-square', 2)
expect_type(ppc.out, "list")
expect_equal(dim(ppc.out$fit.y), c(n.post.samples, N))
expect_equal(dim(ppc.out$fit.y.rep), c(n.post.samples, N))
expect_equal(dim(ppc.out$fit.y.group.quants), c(5, N, max(n.rep)))
expect_equal(dim(ppc.out$fit.y.rep.group.quants), c(5, N, max(n.rep)))
ppc.out <- ppcOcc(out, 'chi-square', 1)
expect_type(ppc.out, "list")
expect_equal(dim(ppc.out$fit.y), c(n.post.samples, N))
expect_equal(dim(ppc.out$fit.y.rep), c(n.post.samples, N))
expect_equal(dim(ppc.out$fit.y.group.quants), c(5, N, J.fit))
expect_equal(dim(ppc.out$fit.y.rep.group.quants), c(5, N, J.fit))
ppc.out <- ppcOcc(out, 'freeman-tukey', 1)
expect_type(ppc.out, "list")
expect_equal(dim(ppc.out$fit.y), c(n.post.samples, N))
expect_equal(dim(ppc.out$fit.y.rep), c(n.post.samples, N))
expect_equal(dim(ppc.out$fit.y.group.quants), c(5, N, J.fit))
expect_equal(dim(ppc.out$fit.y.rep.group.quants), c(5, N, J.fit))
ppc.out <- ppcOcc(out, 'freeman-tukey', 2)
expect_type(ppc.out, "list")
expect_equal(dim(ppc.out$fit.y), c(n.post.samples, N))
expect_equal(dim(ppc.out$fit.y.rep), c(n.post.samples, N))
expect_equal(dim(ppc.out$fit.y.group.quants), c(5, N, max(n.rep)))
expect_equal(dim(ppc.out$fit.y.rep.group.quants), c(5, N, max(n.rep)))
})
# Random intercepts on both -----------------------------------------------
J.x <- 8
J.y <- 8
J <- J.x * J.y
n.rep<- sample(2:4, size = J, replace = TRUE)
N <- 6
# Community-level covariate effects
# Occurrence
beta.mean <- c(0.2)
p.occ <- length(beta.mean)
tau.sq.beta <- c(0.6)
# Detection
alpha.mean <- c(0)
tau.sq.alpha <- c(1)
p.det <- length(alpha.mean)
# Random effects
psi.RE <- list(levels = c(20),
sigma.sq.psi = c(2.5))
p.RE <- list(levels = c(50, 10),
sigma.sq.p = c(2.50, 1.5))
# Draw species-level effects from community means.
beta <- matrix(NA, nrow = N, ncol = p.occ)
alpha <- matrix(NA, nrow = N, ncol = p.det)
for (i in 1:p.occ) {
beta[, i] <- rnorm(N, beta.mean[i], sqrt(tau.sq.beta[i]))
}
for (i in 1:p.det) {
alpha[, i] <- rnorm(N, alpha.mean[i], sqrt(tau.sq.alpha[i]))
}
alpha.true <- alpha
phi <- rep(3 / .7, N)
sigma.sq <- rep(2, N)
nu <- rep(2, N)
dat <- simMsOcc(J.x = J.x, J.y = J.y, n.rep = n.rep, N = N, beta = beta, alpha = alpha,
psi.RE = psi.RE, p.RE = p.RE, sp = TRUE, sigma.sq = sigma.sq,
phi = phi, nu = nu, cov.model = 'matern')
pred.indx <- sample(1:J, round(J * .25), replace = FALSE)
y <- dat$y[, -pred.indx, , drop = FALSE]
# Occupancy covariates
X <- dat$X[-pred.indx, , drop = FALSE]
X.re <- dat$X.re[-pred.indx, , drop = FALSE]
coords <- as.matrix(dat$coords[-pred.indx, , drop = FALSE])
# Prediction covariates
X.0 <- dat$X[pred.indx, , drop = FALSE]
X.re.0 <- dat$X.re[pred.indx, , drop = FALSE]
coords.0 <- as.matrix(dat$coords[pred.indx, , drop = FALSE])
# Detection covariates
X.p <- dat$X.p[-pred.indx, , , drop = FALSE]
X.p.re <- dat$X.p.re[-pred.indx, , , drop = FALSE]
occ.covs <- cbind(X, X.re)
colnames(occ.covs) <- c('int', 'occ.factor.1')
det.covs <- list(det.factor.1 = X.p.re[, , 1],
det.factor.2 = X.p.re[, , 2])
data.list <- list(y = y, coords = coords, occ.covs = occ.covs, det.covs = det.covs)
# Priors
prior.list <- list(beta.comm.normal = list(mean = 0, var = 2.72),
alpha.comm.normal = list(mean = 0, var = 2.72),
tau.sq.beta.ig = list(a = 0.1, b = 0.1),
tau.sq.alpha.ig = list(a = 0.1, b = 0.1),
nu.unif = list(0.5, 2.5))
# Starting values
inits.list <- list(alpha.comm = 0,
beta.comm = 0,
beta = 0,
alpha = 0,
tau.sq.beta = 1,
tau.sq.alpha = 1,
z = apply(y, c(1, 2), max, na.rm = TRUE))
# Tuning
tuning.list <- list(phi = 1, nu = 0.25)
batch.length <- 25
n.batch <- 10
n.report <- 100
occ.formula <- ~ (1 | occ.factor.1)
det.formula <- ~ (1 | det.factor.1) + (1 | det.factor.2)
out <- spMsPGOcc(occ.formula = occ.formula,
det.formula = det.formula,
data = data.list,
inits = inits.list,
n.batch = n.batch,
batch.length = batch.length,
accept.rate = 0.43,
priors = prior.list,
cov.model = "matern",
tuning = tuning.list,
n.omp.threads = 1,
verbose = FALSE,
NNGP = TRUE,
n.neighbors = 5,
search.type = 'cb',
n.report = 10,
n.burn = 100,
n.thin = 2,
n.chains = 2,
k.fold = 2,
k.fold.threads = 1)
# To make sure it worked --------------
test_that("out is of class spMsPGOcc", {
expect_s3_class(out, "spMsPGOcc")
})
# Check cross-validation --------------
test_that("cross-validation works", {
expect_equal(length(out$k.fold.deviance), N)
expect_type(out$k.fold.deviance, "double")
expect_equal(sum(out$k.fold.deviance < 0), 0)
})
# Check random effects ----------------
test_that("random effects are correct", {
expect_equal(out$pRE, TRUE)
expect_equal(out$psiRE, TRUE)
})
# Check RE error ----------------------
test_that("random effect gives error when non-numeric", {
data.list$occ.covs <- as.data.frame(data.list$occ.covs)
data.list$occ.covs$occ.factor.1 <- factor(data.list$occ.covs$occ.factor.1)
data.list$det.covs$det.factor.1 <- factor(data.list$det.covs$det.factor.1)
expect_error(out <- spMsPGOcc(occ.formula = occ.formula,
det.formula = det.formula,
data = data.list,
inits = inits.list,
n.batch = n.batch,
batch.length = batch.length,
accept.rate = 0.43,
priors = prior.list,
cov.model = "matern",
tuning = tuning.list,
n.omp.threads = 1,
verbose = FALSE,
NNGP = TRUE,
n.neighbors = 5,
search.type = 'cb',
n.report = 10,
n.burn = 100,
n.thin = 2,
n.chains = 1))
data.list$occ.covs$occ.factor.1 <- as.character(factor(data.list$occ.covs$occ.factor.1))
data.list$det.covs$det.factor.1 <- as.character(factor(data.list$det.covs$det.factor.1))
expect_error(out <- spMsPGOcc(occ.formula = occ.formula,
det.formula = det.formula,
data = data.list,
inits = inits.list,
n.batch = n.batch,
batch.length = batch.length,
accept.rate = 0.43,
priors = prior.list,
cov.model = "matern",
tuning = tuning.list,
n.omp.threads = 1,
verbose = FALSE,
NNGP = TRUE,
n.neighbors = 5,
search.type = 'cb',
n.report = 10,
n.burn = 100,
n.thin = 2,
n.chains = 1))
})
# Check RE levels ---------------------
test_that("random effect levels are correct", {
expect_equal(sort(unique(unlist(out$re.level.names))),
sort(unique(c(X.re))))
expect_equal(sort(unique(unlist(out$p.re.level.names))),
sort(unique(c(X.p.re))))
})
# Check output data output is correct -
test_that("out$y == y", {
expect_equal(out$y, y)
})
# Check missing values ----------------
test_that("missing value error handling works", {
tmp.data <- data.list
tmp.data$occ.covs[3, ] <- NA
expect_error(spMsPGOcc(occ.formula = occ.formula,
det.formula = det.formula,
data = tmp.data,
n.batch = 40,
batch.length = batch.length,
cov.model = "exponential",
tuning = tuning.list,
NNGP = TRUE,
verbose = FALSE,
n.neighbors = 5,
search.type = 'cb',
n.report = 10,
n.burn = 100,
n.chains = 1))
tmp.data <- data.list
tmp.data$y[1, 1, 1] <- 1
tmp.data$det.covs[[1]][1] <- NA
expect_error(spMsPGOcc(occ.formula = occ.formula,
det.formula = det.formula,
data = tmp.data,
n.batch = 40,
batch.length = batch.length,
cov.model = "exponential",
tuning = tuning.list,
NNGP = TRUE,
verbose = FALSE,
n.neighbors = 5,
search.type = 'cb',
n.report = 10,
n.burn = 100,
n.chains = 1))
tmp.data <- data.list
tmp.data$y[, 1, 1] <- NA
out <- spMsPGOcc(occ.formula = occ.formula,
det.formula = det.formula,
data = tmp.data,
n.batch = 40,
batch.length = batch.length,
cov.model = "exponential",
tuning = tuning.list,
NNGP = TRUE,
verbose = FALSE,
n.neighbors = 5,
search.type = 'cb',
n.report = 10,
n.burn = 100,
n.chains = 1)
expect_s3_class(out, "spMsPGOcc")
})
# Check default values ----------------
test_that("default priors, inits, burn, thin work", {
out <- spMsPGOcc(occ.formula = occ.formula,
det.formula = det.formula,
data = data.list,
n.batch = n.batch,
batch.length = batch.length,
accept.rate = 0.43,
cov.model = "exponential",
tuning = tuning.list,
NNGP = TRUE,
verbose = FALSE,
n.neighbors = 5,
n.chains = 1)
expect_s3_class(out, "spMsPGOcc")
})
test_that("all correlation functions work", {
out <- spMsPGOcc(occ.formula = occ.formula,
det.formula = det.formula,
data = data.list,
inits = inits.list,
n.batch = n.batch,
batch.length = batch.length,
accept.rate = 0.43,
priors = prior.list,
cov.model = "gaussian",
tuning = list(phi = 0.3),
n.omp.threads = 1,
verbose = FALSE,
NNGP = TRUE,
n.neighbors = 5,
search.type = 'cb',
n.report = 10,
n.burn = 100,
n.thin = 1,
n.chains = 1)
expect_s3_class(out, "spMsPGOcc")
out <- spMsPGOcc(occ.formula = occ.formula,
det.formula = det.formula,
data = data.list,
inits = inits.list,
n.batch = n.batch,
batch.length = batch.length,
accept.rate = 0.43,
priors = prior.list,
cov.model = "spherical",
tuning = list(phi = 0.3),
n.omp.threads = 1,
verbose = FALSE,
NNGP = TRUE,
n.neighbors = 5,
search.type = 'cb',
n.report = 10,
n.burn = 100,
n.thin = 1,
n.chains = 1)
expect_s3_class(out, "spMsPGOcc")
out <- spMsPGOcc(occ.formula = occ.formula,
det.formula = det.formula,
data = data.list,
inits = inits.list,
n.batch = n.batch,
batch.length = batch.length,
accept.rate = 0.43,
priors = list(nu.unif = list(0.4, 3)),
cov.model = "matern",
tuning = list(phi = 0.3, nu = 0.2),
n.omp.threads = 1,
verbose = FALSE,
NNGP = TRUE,
n.neighbors = 5,
search.type = 'cb',
n.report = 10,
n.burn = 100,
n.thin = 1,
n.chains = 1)
expect_s3_class(out, "spMsPGOcc")
})
test_that("verbose prints to the screen", {
expect_output(spMsPGOcc(occ.formula = occ.formula,
det.formula = det.formula,
data = data.list,
inits = inits.list,
n.batch = n.batch,
batch.length = batch.length,
accept.rate = 0.43,
priors = prior.list,
cov.model = "exponential",
tuning = tuning.list,
n.omp.threads = 1,
verbose = TRUE,
NNGP = TRUE,
n.neighbors = 5,
search.type = 'cb',
n.report = 10,
n.burn = 100,
n.thin = 1,
n.chains = 1))
})
# Check waicOcc -----------------------
test_that("waicOCC works for spMsPGOcc", {
# as.vector gets rid of names
waic.out <- as.vector(waicOcc(out))
expect_equal(length(waic.out), 3)
expect_equal(waic.out[3], -2 * (waic.out[1] - waic.out[2]))
})
test_that("fitted works for spMsPGOcc", {
fitted.out <- fitted(out)
expect_equal(length(fitted.out), 2)
expect_equal(class(fitted.out$y.rep.samples), "array")
expect_equal(class(fitted.out$p.samples), "array")
expect_equal(dim(fitted.out$y.rep.samples), dim(fitted.out$p.samples))
})
test_that("predict works for spMsPGOcc", {
n.post.samples <- out$n.post * out$n.chains
X.0 <- cbind(X.0, X.re.0)
colnames(X.0) <- c('int', 'occ.factor.1')
pred.out <- predict(out, X.0, coords.0, verbose = FALSE)
expect_type(pred.out, "list")
expect_equal(dim(pred.out$psi.0.samples), c(n.post.samples, N, nrow(X.0)))
expect_equal(dim(pred.out$z.0.samples), c(n.post.samples, N, nrow(X.0)))
})
test_that("detection prediction works", {
X.p.0 <- cbind(dat$X.p[, 1, ], dat$X.p.re[, 1, 1:2])
colnames(X.p.0) <- c('intercept', 'det.factor.1', 'det.factor.2')
pred.out <- predict(out, X.p.0, type = 'detection')
expect_type(pred.out, 'list')
expect_equal(dim(pred.out$p.0.samples), c(out$n.post * out$n.chains, N, J))
})
test_that("posterior predictive checks work for msPGOcc", {
n.post.samples <- out$n.post * out$n.chains
J.fit <- nrow(X)
ppc.out <- ppcOcc(out, 'chi-square', 2)
expect_type(ppc.out, "list")
expect_equal(dim(ppc.out$fit.y), c(n.post.samples, N))
expect_equal(dim(ppc.out$fit.y.rep), c(n.post.samples, N))
expect_equal(dim(ppc.out$fit.y.group.quants), c(5, N, max(n.rep)))
expect_equal(dim(ppc.out$fit.y.rep.group.quants), c(5, N, max(n.rep)))
ppc.out <- ppcOcc(out, 'chi-square', 1)
expect_type(ppc.out, "list")
expect_equal(dim(ppc.out$fit.y), c(n.post.samples, N))
expect_equal(dim(ppc.out$fit.y.rep), c(n.post.samples, N))
expect_equal(dim(ppc.out$fit.y.group.quants), c(5, N, J.fit))
expect_equal(dim(ppc.out$fit.y.rep.group.quants), c(5, N, J.fit))
ppc.out <- ppcOcc(out, 'freeman-tukey', 1)
expect_type(ppc.out, "list")
expect_equal(dim(ppc.out$fit.y), c(n.post.samples, N))
expect_equal(dim(ppc.out$fit.y.rep), c(n.post.samples, N))
expect_equal(dim(ppc.out$fit.y.group.quants), c(5, N, J.fit))
expect_equal(dim(ppc.out$fit.y.rep.group.quants), c(5, N, J.fit))
ppc.out <- ppcOcc(out, 'freeman-tukey', 2)
expect_type(ppc.out, "list")
expect_equal(dim(ppc.out$fit.y), c(n.post.samples, N))
expect_equal(dim(ppc.out$fit.y.rep), c(n.post.samples, N))
expect_equal(dim(ppc.out$fit.y.group.quants), c(5, N, max(n.rep)))
expect_equal(dim(ppc.out$fit.y.rep.group.quants), c(5, N, max(n.rep)))
})
# Random intercepts on both with covariates -------------------------------
J.x <- 8
J.y <- 8
J <- J.x * J.y
n.rep<- sample(2:4, size = J, replace = TRUE)
N <- 6
# Community-level covariate effects
# Occurrence
beta.mean <- c(0.2, 1.2, -0.5)
p.occ <- length(beta.mean)
tau.sq.beta <- c(0.6, 2.4, 0.3)
# Detection
alpha.mean <- c(0, 0.5)
tau.sq.alpha <- c(1, 3.3)
p.det <- length(alpha.mean)
# Random effects
psi.RE <- list(levels = c(20),
sigma.sq.psi = c(2.5))
p.RE <- list(levels = c(50, 10),
sigma.sq.p = c(2.50, 1.5))
# Draw species-level effects from community means.
beta <- matrix(NA, nrow = N, ncol = p.occ)
alpha <- matrix(NA, nrow = N, ncol = p.det)
for (i in 1:p.occ) {
beta[, i] <- rnorm(N, beta.mean[i], sqrt(tau.sq.beta[i]))
}
for (i in 1:p.det) {
alpha[, i] <- rnorm(N, alpha.mean[i], sqrt(tau.sq.alpha[i]))
}
alpha.true <- alpha
phi <- rep(3 / .7, N)
sigma.sq <- rep(2, N)
nu <- rep(2, N)
dat <- simMsOcc(J.x = J.x, J.y = J.y, n.rep = n.rep, N = N, beta = beta, alpha = alpha,
psi.RE = psi.RE, p.RE = p.RE, sp = TRUE, sigma.sq = sigma.sq,
phi = phi, nu = nu, cov.model = 'matern')
pred.indx <- sample(1:J, round(J * .25), replace = FALSE)
y <- dat$y[, -pred.indx, , drop = FALSE]
# Occupancy covariates
X <- dat$X[-pred.indx, , drop = FALSE]
X.re <- dat$X.re[-pred.indx, , drop = FALSE]
coords <- as.matrix(dat$coords[-pred.indx, , drop = FALSE])
# Prediction covariates
X.0 <- dat$X[pred.indx, , drop = FALSE]
X.re.0 <- dat$X.re[pred.indx, , drop = FALSE]
coords.0 <- as.matrix(dat$coords[pred.indx, , drop = FALSE])
# Detection covariates
X.p <- dat$X.p[-pred.indx, , , drop = FALSE]
X.p.re <- dat$X.p.re[-pred.indx, , , drop = FALSE]
occ.covs <- cbind(X, X.re)
colnames(occ.covs) <- c('int', 'occ.cov.1', 'occ.cov.2', 'occ.factor.1')
det.covs <- list(det.cov.1 = X.p[, , 2],
det.factor.1 = X.p.re[, , 1],
det.factor.2 = X.p.re[, , 2])
data.list <- list(y = y, coords = coords, occ.covs = occ.covs, det.covs = det.covs)
# Priors
prior.list <- list(beta.comm.normal = list(mean = 0, var = 2.72),
alpha.comm.normal = list(mean = 0, var = 2.72),
tau.sq.beta.ig = list(a = 0.1, b = 0.1),
tau.sq.alpha.ig = list(a = 0.1, b = 0.1),
nu.unif = list(0.5, 2.5))
# Starting values
inits.list <- list(alpha.comm = 0,
beta.comm = 0,
beta = 0,
alpha = 0,
tau.sq.beta = 1,
tau.sq.alpha = 1,
z = apply(y, c(1, 2), max, na.rm = TRUE))
# Tuning
tuning.list <- list(phi = 1, nu = 0.25)
batch.length <- 25
n.batch <- 10
n.report <- 100
occ.formula <- ~ occ.cov.1 + occ.cov.2 + (1 | occ.factor.1)
det.formula <- ~ det.cov.1 + (1 | det.factor.1) + (1 | det.factor.2)
out <- spMsPGOcc(occ.formula = occ.formula,
det.formula = det.formula,
data = data.list,
inits = inits.list,
n.batch = n.batch,
batch.length = batch.length,
accept.rate = 0.43,
priors = prior.list,
cov.model = "matern",
tuning = tuning.list,
n.omp.threads = 1,
verbose = FALSE,
NNGP = TRUE,
n.neighbors = 5,
search.type = 'cb',
n.report = 10,
n.burn = 100,
n.thin = 2,
n.chains = 2,
k.fold = 2,
k.fold.threads = 1)
# To make sure it worked --------------
test_that("out is of class spMsPGOcc", {
expect_s3_class(out, "spMsPGOcc")
})
# Check cross-validation --------------
test_that("cross-validation works", {
expect_equal(length(out$k.fold.deviance), N)
expect_type(out$k.fold.deviance, "double")
expect_equal(sum(out$k.fold.deviance < 0), 0)
})
# Check random effects ----------------
test_that("random effects are correct", {
expect_equal(out$pRE, TRUE)
expect_equal(out$psiRE, TRUE)
})
# Check RE error ----------------------
test_that("random effect gives error when non-numeric", {
data.list$occ.covs <- as.data.frame(data.list$occ.covs)
data.list$occ.covs$occ.factor.1 <- factor(data.list$occ.covs$occ.factor.1)
data.list$det.covs$det.factor.1 <- factor(data.list$det.covs$det.factor.1)
expect_error(out <- spMsPGOcc(occ.formula = occ.formula,
det.formula = det.formula,
data = data.list,
inits = inits.list,
n.batch = n.batch,
batch.length = batch.length,
accept.rate = 0.43,
priors = prior.list,
cov.model = "matern",
tuning = tuning.list,
n.omp.threads = 1,
verbose = FALSE,
NNGP = TRUE,
n.neighbors = 5,
search.type = 'cb',
n.report = 10,
n.burn = 100,
n.thin = 2,
n.chains = 1))
data.list$occ.covs$occ.factor.1 <- as.character(factor(data.list$occ.covs$occ.factor.1))
data.list$det.covs$det.factor.1 <- as.character(factor(data.list$det.covs$det.factor.1))
expect_error(out <- spMsPGOcc(occ.formula = occ.formula,
det.formula = det.formula,
data = data.list,
inits = inits.list,
n.batch = n.batch,
batch.length = batch.length,
accept.rate = 0.43,
priors = prior.list,
cov.model = "matern",
tuning = tuning.list,
n.omp.threads = 1,
verbose = FALSE,
NNGP = TRUE,
n.neighbors = 5,
search.type = 'cb',
n.report = 10,
n.burn = 100,
n.thin = 2,
n.chains = 1))
})
# Check RE levels ---------------------
test_that("random effect levels are correct", {
expect_equal(sort(unique(unlist(out$re.level.names))),
sort(unique(c(X.re))))
expect_equal(sort(unique(unlist(out$p.re.level.names))),
sort(unique(c(X.p.re))))
})
# Check output data output is correct -
test_that("out$y == y", {
expect_equal(out$y, y)
})
# Check missing values ----------------
test_that("missing value error handling works", {
tmp.data <- data.list
tmp.data$occ.covs[3, ] <- NA
expect_error(spMsPGOcc(occ.formula = occ.formula,
det.formula = det.formula,
data = tmp.data,
n.batch = 40,
batch.length = batch.length,
cov.model = "exponential",
tuning = tuning.list,
NNGP = TRUE,
verbose = FALSE,
n.neighbors = 5,
search.type = 'cb',
n.report = 10,
n.burn = 100,
n.chains = 1))
tmp.data <- data.list
tmp.data$y[1, 1, 1] <- 1
tmp.data$det.covs[[1]][1] <- NA
expect_error(spMsPGOcc(occ.formula = occ.formula,
det.formula = det.formula,
data = tmp.data,
n.batch = 40,
batch.length = batch.length,
cov.model = "exponential",
tuning = tuning.list,
NNGP = TRUE,
verbose = FALSE,
n.neighbors = 5,
search.type = 'cb',
n.report = 10,
n.burn = 100,
n.chains = 1))
tmp.data <- data.list
tmp.data$y[, 1, 1] <- NA
out <- spMsPGOcc(occ.formula = occ.formula,
det.formula = det.formula,
data = tmp.data,
n.batch = 40,
batch.length = batch.length,
cov.model = "exponential",
tuning = tuning.list,
NNGP = TRUE,
verbose = FALSE,
n.neighbors = 5,
search.type = 'cb',
n.report = 10,
n.burn = 100,
n.chains = 1)
expect_s3_class(out, "spMsPGOcc")
})
# Check default values ----------------
test_that("default priors, inits, burn, thin work", {
out <- spMsPGOcc(occ.formula = occ.formula,
det.formula = det.formula,
data = data.list,
n.batch = n.batch,
batch.length = batch.length,
accept.rate = 0.43,
cov.model = "exponential",
tuning = tuning.list,
NNGP = TRUE,
verbose = FALSE,
n.neighbors = 5,
n.chains = 1)
expect_s3_class(out, "spMsPGOcc")
})
test_that("all correlation functions work", {
out <- spMsPGOcc(occ.formula = occ.formula,
det.formula = det.formula,
data = data.list,
inits = inits.list,
n.batch = n.batch,
batch.length = batch.length,
accept.rate = 0.43,
priors = prior.list,
cov.model = "gaussian",
tuning = list(phi = 0.3),
n.omp.threads = 1,
verbose = FALSE,
NNGP = TRUE,
n.neighbors = 5,
search.type = 'cb',
n.report = 10,
n.burn = 100,
n.thin = 1,
n.chains = 1)
expect_s3_class(out, "spMsPGOcc")
out <- spMsPGOcc(occ.formula = occ.formula,
det.formula = det.formula,
data = data.list,
inits = inits.list,
n.batch = n.batch,
batch.length = batch.length,
accept.rate = 0.43,
priors = prior.list,
cov.model = "spherical",
tuning = list(phi = 0.3),
n.omp.threads = 1,
verbose = FALSE,
NNGP = TRUE,
n.neighbors = 5,
search.type = 'cb',
n.report = 10,
n.burn = 100,
n.thin = 1,
n.chains = 1)
expect_s3_class(out, "spMsPGOcc")
out <- spMsPGOcc(occ.formula = occ.formula,
det.formula = det.formula,
data = data.list,
inits = inits.list,
n.batch = n.batch,
batch.length = batch.length,
accept.rate = 0.43,
priors = list(nu.unif = list(0.4, 3)),
cov.model = "matern",
tuning = list(phi = 0.3, nu = 0.2),
n.omp.threads = 1,
verbose = FALSE,
NNGP = TRUE,
n.neighbors = 5,
search.type = 'cb',
n.report = 10,
n.burn = 100,
n.thin = 1,
n.chains = 1)
expect_s3_class(out, "spMsPGOcc")
})
test_that("verbose prints to the screen", {
expect_output(spMsPGOcc(occ.formula = occ.formula,
det.formula = det.formula,
data = data.list,
inits = inits.list,
n.batch = n.batch,
batch.length = batch.length,
accept.rate = 0.43,
priors = prior.list,
cov.model = "exponential",
tuning = tuning.list,
n.omp.threads = 1,
verbose = TRUE,
NNGP = TRUE,
n.neighbors = 5,
search.type = 'cb',
n.report = 10,
n.burn = 100,
n.thin = 1,
n.chains = 1))
})
# Check waicOcc -----------------------
test_that("waicOCC works for spMsPGOcc", {
# as.vector gets rid of names
waic.out <- as.vector(waicOcc(out))
expect_equal(length(waic.out), 3)
expect_equal(waic.out[3], -2 * (waic.out[1] - waic.out[2]))
})
test_that("fitted works for spMsPGOcc", {
fitted.out <- fitted(out)
expect_equal(length(fitted.out), 2)
expect_equal(class(fitted.out$y.rep.samples), "array")
expect_equal(class(fitted.out$p.samples), "array")
expect_equal(dim(fitted.out$y.rep.samples), dim(fitted.out$p.samples))
})
test_that("predict works for spMsPGOcc", {
n.post.samples <- out$n.post * out$n.chains
X.0 <- cbind(X.0, X.re.0)
colnames(X.0) <- c('int', 'occ.cov.1', 'occ.cov.2', 'occ.factor.1')
pred.out <- predict(out, X.0, coords.0, verbose = FALSE)
expect_type(pred.out, "list")
expect_equal(dim(pred.out$psi.0.samples), c(n.post.samples, N, nrow(X.0)))
expect_equal(dim(pred.out$z.0.samples), c(n.post.samples, N, nrow(X.0)))
})
test_that("detection prediction works", {
X.p.0 <- cbind(dat$X.p[, 1, ], dat$X.p.re[, 1, 1:2])
colnames(X.p.0) <- c('intercept', 'det.cov.1', 'det.factor.1', 'det.factor.2')
pred.out <- predict(out, X.p.0, type = 'detection')
expect_type(pred.out, 'list')
expect_equal(dim(pred.out$p.0.samples), c(out$n.post * out$n.chains, N, J))
})
test_that("posterior predictive checks work for msPGOcc", {
n.post.samples <- out$n.post * out$n.chains
J.fit <- nrow(X)
ppc.out <- ppcOcc(out, 'chi-square', 2)
expect_type(ppc.out, "list")
expect_equal(dim(ppc.out$fit.y), c(n.post.samples, N))
expect_equal(dim(ppc.out$fit.y.rep), c(n.post.samples, N))
expect_equal(dim(ppc.out$fit.y.group.quants), c(5, N, max(n.rep)))
expect_equal(dim(ppc.out$fit.y.rep.group.quants), c(5, N, max(n.rep)))
ppc.out <- ppcOcc(out, 'chi-square', 1)
expect_type(ppc.out, "list")
expect_equal(dim(ppc.out$fit.y), c(n.post.samples, N))
expect_equal(dim(ppc.out$fit.y.rep), c(n.post.samples, N))
expect_equal(dim(ppc.out$fit.y.group.quants), c(5, N, J.fit))
expect_equal(dim(ppc.out$fit.y.rep.group.quants), c(5, N, J.fit))
ppc.out <- ppcOcc(out, 'freeman-tukey', 1)
expect_type(ppc.out, "list")
expect_equal(dim(ppc.out$fit.y), c(n.post.samples, N))
expect_equal(dim(ppc.out$fit.y.rep), c(n.post.samples, N))
expect_equal(dim(ppc.out$fit.y.group.quants), c(5, N, J.fit))
expect_equal(dim(ppc.out$fit.y.rep.group.quants), c(5, N, J.fit))
ppc.out <- ppcOcc(out, 'freeman-tukey', 2)
expect_type(ppc.out, "list")
expect_equal(dim(ppc.out$fit.y), c(n.post.samples, N))
expect_equal(dim(ppc.out$fit.y.rep), c(n.post.samples, N))
expect_equal(dim(ppc.out$fit.y.group.quants), c(5, N, max(n.rep)))
expect_equal(dim(ppc.out$fit.y.rep.group.quants), c(5, N, max(n.rep)))
})
# Model with fixed sigma.sq -----------------------------------------------
test_that("spMsPGOcc works with fixed sigma.sq", {
out <- spMsPGOcc(occ.formula = occ.formula,
det.formula = det.formula,
data = data.list,
inits = inits.list,
n.batch = n.batch,
batch.length = batch.length,
accept.rate = 0.43,
priors = list(sigma.sq.ig = "fixed"),
cov.model = "exponential",
tuning = list(phi = 0.3, nu = 0.2),
n.omp.threads = 1,
verbose = FALSE,
NNGP = TRUE,
n.neighbors = 5,
search.type = 'cb',
n.report = 10,
n.burn = 100,
n.thin = 1,
n.chains = 1)
expect_s3_class(out, "spMsPGOcc")
expect_equal(length(unique(out$theta.samples[, 1])), 1)
})
# Uniform sigma sq --------------------------------------------------------
test_that("spMsPGOcc works with uniform prior on sigma.sq", {
prior.list <- list(sigma.sq.unif = list(a = 0, b = 5),
nu.unif = list(a = 0.1, b = 4))
tuning.list <- list(phi = 0.5, nu = 0.6, sigma.sq = 0.7)
out <- spMsPGOcc(occ.formula = occ.formula,
det.formula = det.formula,
data = data.list,
n.batch = 40,
batch.length = batch.length,
cov.model = "exponential",
priors = prior.list,
tuning = tuning.list,
NNGP = TRUE,
verbose = FALSE,
n.neighbors = 5,
search.type = 'cb',
n.report = 10,
n.burn = 100,
n.chains = 1)
expect_s3_class(out, "spMsPGOcc")
out <- spMsPGOcc(occ.formula = occ.formula,
det.formula = det.formula,
data = data.list,
n.batch = 40,
batch.length = batch.length,
cov.model = "matern",
priors = prior.list,
tuning = tuning.list,
NNGP = TRUE,
verbose = FALSE,
n.neighbors = 5,
search.type = 'cb',
n.report = 10,
n.burn = 100,
n.chains = 1)
expect_s3_class(out, "spMsPGOcc")
})
# Third dimension of y != max(n.rep)
J.x <- 8
J.y <- 8
J <- J.x * J.y
n.rep<- sample(2:4, size = J, replace = TRUE)
n.rep.max <- 7
N <- 6
# Community-level covariate effects
# Occurrence
beta.mean <- c(0.2, 0.5, 1.2)
p.occ <- length(beta.mean)
tau.sq.beta <- c(0.6, 1.5, 2.3)
# Detection
alpha.mean <- c(0, -0.5)
tau.sq.alpha <- c(1, 2.3)
p.det <- length(alpha.mean)
# Random effects
psi.RE <- list()
p.RE <- list()
# Draw species-level effects from community means.
beta <- matrix(NA, nrow = N, ncol = p.occ)
alpha <- matrix(NA, nrow = N, ncol = p.det)
for (i in 1:p.occ) {
beta[, i] <- rnorm(N, beta.mean[i], sqrt(tau.sq.beta[i]))
}
for (i in 1:p.det) {
alpha[, i] <- rnorm(N, alpha.mean[i], sqrt(tau.sq.alpha[i]))
}
alpha.true <- alpha
phi <- rep(3 / .7, N)
sigma.sq <- rep(2, N)
nu <- rep(2, N)
dat <- simMsOcc(J.x = J.x, J.y = J.y, n.rep = n.rep, N = N, beta = beta, alpha = alpha,
psi.RE = psi.RE, p.RE = p.RE, sp = TRUE, sigma.sq = sigma.sq,
phi = phi, nu = nu, cov.model = 'matern', n.rep.max = n.rep.max)
pred.indx <- sample(1:J, round(J * .25), replace = FALSE)
y <- dat$y[, -pred.indx, , drop = FALSE]
# Occupancy covariates
X <- dat$X[-pred.indx, , drop = FALSE]
coords <- as.matrix(dat$coords[-pred.indx, , drop = FALSE])
# Prediction covariates
X.0 <- dat$X[pred.indx, , drop = FALSE]
coords.0 <- as.matrix(dat$coords[pred.indx, , drop = FALSE])
# Detection covariates
X.p <- dat$X.p[-pred.indx, , , drop = FALSE]
occ.covs <- X
colnames(occ.covs) <- c('int', 'occ.cov.1', 'occ.cov.2')
det.covs <- list(det.cov.1 = X.p[, , 2])
data.list <- list(y = y, coords = coords, occ.covs = occ.covs, det.covs = det.covs)
# Priors
prior.list <- list(beta.comm.normal = list(mean = 0, var = 2.72),
alpha.comm.normal = list(mean = 0, var = 2.72),
tau.sq.beta.ig = list(a = 0.1, b = 0.1),
tau.sq.alpha.ig = list(a = 0.1, b = 0.1),
nu.unif = list(0.5, 2.5))
# Starting values
inits.list <- list(alpha.comm = 0,
beta.comm = 0,
beta = 0,
alpha = 0,
tau.sq.beta = 1,
tau.sq.alpha = 1,
z = apply(y, c(1, 2), max, na.rm = TRUE))
# Tuning
tuning.list <- list(phi = 1, nu = 0.25)
batch.length <- 25
n.batch <- 10
n.report <- 100
occ.formula <- ~ occ.cov.1 + occ.cov.2
det.formula <- ~ det.cov.1
out <- spMsPGOcc(occ.formula = occ.formula,
det.formula = det.formula,
data = data.list,
inits = inits.list,
n.batch = n.batch,
batch.length = batch.length,
accept.rate = 0.43,
priors = prior.list,
cov.model = "matern",
tuning = tuning.list,
n.omp.threads = 1,
verbose = FALSE,
NNGP = TRUE,
n.neighbors = 5,
search.type = 'cb',
n.report = 10,
n.burn = 100,
n.thin = 2,
n.chains = 2,
k.fold = 2,
k.fold.threads = 1)
# To make sure it worked --------------
test_that("out is of class spMsPGOcc", {
expect_s3_class(out, "spMsPGOcc")
})
# Check cross-validation --------------
test_that("cross-validation works", {
expect_equal(length(out$k.fold.deviance), N)
expect_type(out$k.fold.deviance, "double")
expect_equal(sum(out$k.fold.deviance < 0), 0)
})
# Check random effects ----------------
test_that("random effects are correct", {
expect_equal(out$pRE, FALSE)
expect_equal(out$psiRE, FALSE)
})
# Check output data output is correct -
test_that("out$y == y", {
expect_equal(out$y, y)
})
# Check missing values ----------------
test_that("missing value error handling works", {
tmp.data <- data.list
tmp.data$occ.covs[3, ] <- NA
expect_error(spMsPGOcc(occ.formula = occ.formula,
det.formula = det.formula,
data = tmp.data,
n.batch = 40,
batch.length = batch.length,
cov.model = "exponential",
tuning = tuning.list,
NNGP = TRUE,
verbose = FALSE,
n.neighbors = 5,
search.type = 'cb',
n.report = 10,
n.burn = 100,
n.chains = 1))
tmp.data <- data.list
tmp.data$y[1, 1, 1] <- 1
tmp.data$det.covs[[1]][1] <- NA
expect_error(spMsPGOcc(occ.formula = occ.formula,
det.formula = det.formula,
data = tmp.data,
n.batch = 40,
batch.length = batch.length,
cov.model = "exponential",
tuning = tuning.list,
NNGP = TRUE,
verbose = FALSE,
n.neighbors = 5,
search.type = 'cb',
n.report = 10,
n.burn = 100,
n.chains = 1))
tmp.data <- data.list
tmp.data$y[, 1, 1] <- NA
out <- spMsPGOcc(occ.formula = occ.formula,
det.formula = det.formula,
data = tmp.data,
n.batch = 40,
batch.length = batch.length,
cov.model = "exponential",
tuning = tuning.list,
NNGP = TRUE,
verbose = FALSE,
n.neighbors = 5,
search.type = 'cb',
n.report = 10,
n.burn = 100,
n.chains = 1)
expect_s3_class(out, "spMsPGOcc")
})
# Check default values ----------------
test_that("default priors, inits, burn, thin work", {
out <- spMsPGOcc(occ.formula = occ.formula,
det.formula = det.formula,
data = data.list,
n.batch = n.batch,
batch.length = batch.length,
accept.rate = 0.43,
cov.model = "exponential",
tuning = tuning.list,
NNGP = TRUE,
verbose = FALSE,
n.neighbors = 5,
n.chains = 1)
expect_s3_class(out, "spMsPGOcc")
})
test_that("all correlation functions work", {
out <- spMsPGOcc(occ.formula = occ.formula,
det.formula = det.formula,
data = data.list,
inits = inits.list,
n.batch = n.batch,
batch.length = batch.length,
accept.rate = 0.43,
priors = prior.list,
cov.model = "gaussian",
tuning = list(phi = 0.3),
n.omp.threads = 1,
verbose = FALSE,
NNGP = TRUE,
n.neighbors = 5,
search.type = 'cb',
n.report = 10,
n.burn = 100,
n.thin = 1,
n.chains = 1)
expect_s3_class(out, "spMsPGOcc")
out <- spMsPGOcc(occ.formula = occ.formula,
det.formula = det.formula,
data = data.list,
inits = inits.list,
n.batch = n.batch,
batch.length = batch.length,
accept.rate = 0.43,
priors = prior.list,
cov.model = "spherical",
tuning = list(phi = 0.3),
n.omp.threads = 1,
verbose = FALSE,
NNGP = TRUE,
n.neighbors = 5,
search.type = 'cb',
n.report = 10,
n.burn = 100,
n.thin = 1,
n.chains = 1)
expect_s3_class(out, "spMsPGOcc")
out <- spMsPGOcc(occ.formula = occ.formula,
det.formula = det.formula,
data = data.list,
inits = inits.list,
n.batch = n.batch,
batch.length = batch.length,
accept.rate = 0.43,
priors = list(nu.unif = list(0.4, 3)),
cov.model = "matern",
tuning = list(phi = 0.3, nu = 0.2),
n.omp.threads = 1,
verbose = FALSE,
NNGP = TRUE,
n.neighbors = 5,
search.type = 'cb',
n.report = 10,
n.burn = 100,
n.thin = 1,
n.chains = 1)
expect_s3_class(out, "spMsPGOcc")
})
test_that("verbose prints to the screen", {
expect_output(spMsPGOcc(occ.formula = occ.formula,
det.formula = det.formula,
data = data.list,
inits = inits.list,
n.batch = n.batch,
batch.length = batch.length,
accept.rate = 0.43,
priors = prior.list,
cov.model = "exponential",
tuning = tuning.list,
n.omp.threads = 1,
verbose = TRUE,
NNGP = TRUE,
n.neighbors = 5,
search.type = 'cb',
n.report = 10,
n.burn = 100,
n.thin = 1,
n.chains = 1))
})
# Check waicOcc -----------------------
test_that("waicOCC works for spMsPGOcc", {
# as.vector gets rid of names
waic.out <- as.vector(waicOcc(out))
expect_equal(length(waic.out), 3)
expect_equal(waic.out[3], -2 * (waic.out[1] - waic.out[2]))
})
test_that("fitted works for spMsPGOcc", {
fitted.out <- fitted(out)
expect_equal(length(fitted.out), 2)
expect_equal(class(fitted.out$y.rep.samples), "array")
expect_equal(class(fitted.out$p.samples), "array")
expect_equal(dim(fitted.out$y.rep.samples), dim(fitted.out$p.samples))
})
test_that("predict works for spMsPGOcc", {
n.post.samples <- out$n.post * out$n.chains
pred.out <- predict(out, X.0, coords.0, verbose = FALSE)
expect_type(pred.out, "list")
expect_equal(dim(pred.out$psi.0.samples), c(n.post.samples, N, nrow(X.0)))
expect_equal(dim(pred.out$z.0.samples), c(n.post.samples, N, nrow(X.0)))
})
test_that("detection prediction works", {
X.p.0 <- dat$X.p[, 1, ]
pred.out <- predict(out, X.p.0, type = 'detection')
expect_type(pred.out, 'list')
expect_equal(dim(pred.out$p.0.samples), c(out$n.post * out$n.chains, N, J))
})
test_that("posterior predictive checks work for msPGOcc", {
n.post.samples <- out$n.post * out$n.chains
J.fit <- nrow(X)
ppc.out <- ppcOcc(out, 'chi-square', 2)
expect_type(ppc.out, "list")
expect_equal(dim(ppc.out$fit.y), c(n.post.samples, N))
expect_equal(dim(ppc.out$fit.y.rep), c(n.post.samples, N))
expect_equal(dim(ppc.out$fit.y.group.quants), c(5, N, n.rep.max))
expect_equal(dim(ppc.out$fit.y.rep.group.quants), c(5, N, n.rep.max))
ppc.out <- ppcOcc(out, 'chi-square', 1)
expect_type(ppc.out, "list")
expect_equal(dim(ppc.out$fit.y), c(n.post.samples, N))
expect_equal(dim(ppc.out$fit.y.rep), c(n.post.samples, N))
expect_equal(dim(ppc.out$fit.y.group.quants), c(5, N, J.fit))
expect_equal(dim(ppc.out$fit.y.rep.group.quants), c(5, N, J.fit))
ppc.out <- ppcOcc(out, 'freeman-tukey', 1)
expect_type(ppc.out, "list")
expect_equal(dim(ppc.out$fit.y), c(n.post.samples, N))
expect_equal(dim(ppc.out$fit.y.rep), c(n.post.samples, N))
expect_equal(dim(ppc.out$fit.y.group.quants), c(5, N, J.fit))
expect_equal(dim(ppc.out$fit.y.rep.group.quants), c(5, N, J.fit))
ppc.out <- ppcOcc(out, 'freeman-tukey', 2)
expect_type(ppc.out, "list")
expect_equal(dim(ppc.out$fit.y), c(n.post.samples, N))
expect_equal(dim(ppc.out$fit.y.rep), c(n.post.samples, N))
expect_equal(dim(ppc.out$fit.y.group.quants), c(5, N, n.rep.max))
expect_equal(dim(ppc.out$fit.y.rep.group.quants), c(5, N, n.rep.max))
})
doserjef/spOccupancy documentation built on Feb. 28, 2025, 1:19 p.m.
R Package Documentation
Browse R Packages
We want your feedback!
Note that we can't provide technical support on individual packages. You should contact the package authors for that.
# Test spMsPGOcc.R -------------------------------------------------------
# NNGP --------------------------------------------------------------------
skip_on_cran()
# Intercept Only ----------------------------------------------------------
set.seed(1010)
J.x <- 8
J.y <- 8
J <- J.x * J.y
n.rep<- sample(2:4, size = J, replace = TRUE)
N <- 6
# Community-level covariate effects
# Occurrence
beta.mean <- c(0.2)
p.occ <- length(beta.mean)
tau.sq.beta <- c(0.6)
# Detection
alpha.mean <- c(0)
tau.sq.alpha <- c(1)
p.det <- length(alpha.mean)
# Random effects
psi.RE <- list()
p.RE <- list()
# Draw species-level effects from community means.
beta <- matrix(NA, nrow = N, ncol = p.occ)
alpha <- matrix(NA, nrow = N, ncol = p.det)
for (i in 1:p.occ) {
beta[, i] <- rnorm(N, beta.mean[i], sqrt(tau.sq.beta[i]))
}
for (i in 1:p.det) {
alpha[, i] <- rnorm(N, alpha.mean[i], sqrt(tau.sq.alpha[i]))
}
alpha.true <- alpha
phi <- rep(3 / .7, N)
sigma.sq <- rep(2, N)
nu <- rep(2, N)
dat <- simMsOcc(J.x = J.x, J.y = J.y, n.rep = n.rep, N = N, beta = beta, alpha = alpha,
psi.RE = psi.RE, p.RE = p.RE, sp = TRUE, sigma.sq = sigma.sq,
phi = phi, nu = nu, cov.model = 'matern')
pred.indx <- sample(1:J, round(J * .25), replace = FALSE)
y <- dat$y[, -pred.indx, , drop = FALSE]
# Occupancy covariates
X <- dat$X[-pred.indx, , drop = FALSE]
coords <- as.matrix(dat$coords[-pred.indx, , drop = FALSE])
# Prediction covariates
X.0 <- dat$X[pred.indx, , drop = FALSE]
coords.0 <- as.matrix(dat$coords[pred.indx, , drop = FALSE])
# Detection covariates
X.p <- dat$X.p[-pred.indx, , , drop = FALSE]
data.list <- list(y = y, coords = coords)
# Priors
prior.list <- list(beta.comm.normal = list(mean = 0, var = 2.72),
alpha.comm.normal = list(mean = 0, var = 2.72),
tau.sq.beta.ig = list(a = 0.1, b = 0.1),
tau.sq.alpha.ig = list(a = 0.1, b = 0.1),
nu.unif = list(0.5, 2.5))
# Starting values
inits.list <- list(alpha.comm = 0,
beta.comm = 0,
beta = 0,
alpha = 0,
tau.sq.beta = 1,
tau.sq.alpha = 1,
fix = TRUE,
z = apply(y, c(1, 2), max, na.rm = TRUE))
# Tuning
tuning.list <- list(phi = 1, nu = 0.25)
batch.length <- 25
n.batch <- 10
n.report <- 100
occ.formula <- ~ 1
det.formula <- ~ 1
out <- spMsPGOcc(occ.formula = occ.formula,
det.formula = det.formula,
data = data.list,
inits = inits.list,
n.batch = n.batch,
batch.length = batch.length,
accept.rate = 0.43,
priors = prior.list,
cov.model = "matern",
tuning = tuning.list,
n.omp.threads = 1,
verbose = FALSE,
NNGP = TRUE,
n.neighbors = 5,
search.type = 'cb',
n.report = 10,
n.burn = 100,
n.thin = 2,
n.chains = 2,
k.fold = 2,
k.fold.threads = 1)
# To make sure it worked --------------
test_that("out is of class spMsPGOcc", {
expect_s3_class(out, "spMsPGOcc")
})
# Check cross-validation --------------
test_that("cross-validation works", {
out.k.fold <- spMsPGOcc(occ.formula = occ.formula,
det.formula = det.formula,
data = data.list,
inits = inits.list,
n.batch = n.batch,
batch.length = batch.length,
accept.rate = 0.43,
priors = prior.list,
cov.model = "matern",
tuning = tuning.list,
n.omp.threads = 1,
verbose = FALSE,
NNGP = TRUE,
n.neighbors = 5,
search.type = 'cb',
n.report = 10,
n.burn = 100,
n.thin = 2,
n.chains = 2,
k.fold = 2,
k.fold.only = TRUE,
k.fold.threads = 1)
expect_equal(length(out.k.fold$k.fold.deviance), N)
expect_type(out.k.fold$k.fold.deviance, "double")
expect_equal(sum(out.k.fold$k.fold.deviance < 0), 0)
})
# Check random effects ----------------
test_that("random effects are correct", {
expect_equal(out$pRE, FALSE)
expect_equal(out$psiRE, FALSE)
})
# Check output data output is correct -
test_that("out$y == y", {
expect_equal(out$y, y)
})
# Check non-integer n.post -------------
test_that("non-integer n.post", {
expect_error(out <- spMsPGOcc(occ.formula = occ.formula,
det.formula = det.formula,
data = data.list,
n.thin = 13,
n.batch = n.batch,
batch.length = batch.length,
accept.rate = 0.43,
n.omp.threads = 1,
verbose = FALSE))
})
# Check default values ----------------
test_that("default priors, inits, burn, thin work", {
out <- spMsPGOcc(occ.formula = occ.formula,
det.formula = det.formula,
data = data.list,
n.batch = n.batch,
batch.length = batch.length,
accept.rate = 0.43,
cov.model = "exponential",
tuning = tuning.list,
NNGP = TRUE,
verbose = FALSE,
n.neighbors = 5,
n.chains = 1)
expect_s3_class(out, "spMsPGOcc")
})
test_that("all correlation functions work", {
out <- spMsPGOcc(occ.formula = occ.formula,
det.formula = det.formula,
data = data.list,
inits = inits.list,
n.batch = n.batch,
batch.length = batch.length,
accept.rate = 0.43,
priors = prior.list,
cov.model = "gaussian",
tuning = list(phi = 0.3),
n.omp.threads = 1,
verbose = FALSE,
NNGP = TRUE,
n.neighbors = 5,
search.type = 'cb',
n.report = 10,
n.burn = 100,
n.thin = 1,
n.chains = 1)
expect_s3_class(out, "spMsPGOcc")
out <- spMsPGOcc(occ.formula = occ.formula,
det.formula = det.formula,
data = data.list,
inits = inits.list,
n.batch = n.batch,
batch.length = batch.length,
accept.rate = 0.43,
priors = prior.list,
cov.model = "spherical",
tuning = list(phi = 0.3),
n.omp.threads = 1,
verbose = FALSE,
NNGP = TRUE,
n.neighbors = 5,
search.type = 'cb',
n.report = 10,
n.burn = 100,
n.thin = 1,
n.chains = 1)
expect_s3_class(out, "spMsPGOcc")
out <- spMsPGOcc(occ.formula = occ.formula,
det.formula = det.formula,
data = data.list,
inits = inits.list,
n.batch = n.batch,
batch.length = batch.length,
accept.rate = 0.43,
priors = list(nu.unif = list(0.4, 3)),
cov.model = "matern",
tuning = list(phi = 0.3, nu = 0.2),
n.omp.threads = 1,
verbose = FALSE,
NNGP = TRUE,
n.neighbors = 5,
search.type = 'cb',
n.report = 10,
n.burn = 100,
n.thin = 1,
n.chains = 1)
expect_s3_class(out, "spMsPGOcc")
})
test_that("verbose prints to the screen", {
expect_output(spMsPGOcc(occ.formula = occ.formula,
det.formula = det.formula,
data = data.list,
inits = inits.list,
n.batch = n.batch,
batch.length = batch.length,
accept.rate = 0.43,
priors = prior.list,
cov.model = "exponential",
tuning = tuning.list,
n.omp.threads = 1,
verbose = TRUE,
NNGP = TRUE,
n.neighbors = 5,
search.type = 'cb',
n.report = 10,
n.burn = 100,
n.thin = 1,
n.chains = 1))
})
# Check waicOcc -----------------------
test_that("waicOCC works for spMsPGOcc", {
# as.vector gets rid of names
waic.out <- as.vector(waicOcc(out))
expect_equal(length(waic.out), 3)
expect_equal(waic.out[3], -2 * (waic.out[1] - waic.out[2]))
})
test_that("fitted works for spMsPGOcc", {
fitted.out <- fitted(out)
expect_equal(length(fitted.out), 2)
expect_equal(class(fitted.out$y.rep.samples), "array")
expect_equal(class(fitted.out$p.samples), "array")
expect_equal(dim(fitted.out$y.rep.samples), dim(fitted.out$p.samples))
})
test_that("predict works for spMsPGOcc", {
n.post.samples <- out$n.post * out$n.chains
pred.out <- predict(out, X.0, coords.0, verbose = FALSE)
expect_type(pred.out, "list")
expect_equal(dim(pred.out$psi.0.samples), c(n.post.samples, N, nrow(X.0)))
expect_equal(dim(pred.out$z.0.samples), c(n.post.samples, N, nrow(X.0)))
})
test_that("detection prediction works", {
J.str <- 100
X.p.0 <- matrix(1, nrow = J.str, ncol = p.det)
pred.out <- predict(out, X.p.0, type = 'detection')
expect_type(pred.out, 'list')
expect_equal(dim(pred.out$p.0.samples), c(out$n.post * out$n.chains, N, J.str))
})
test_that("posterior predictive checks work for msPGOcc", {
n.post.samples <- out$n.post * out$n.chains
J.fit <- nrow(X)
ppc.out <- ppcOcc(out, 'chi-square', 2)
expect_type(ppc.out, "list")
expect_equal(dim(ppc.out$fit.y), c(n.post.samples, N))
expect_equal(dim(ppc.out$fit.y.rep), c(n.post.samples, N))
expect_equal(dim(ppc.out$fit.y.group.quants), c(5, N, max(n.rep)))
expect_equal(dim(ppc.out$fit.y.rep.group.quants), c(5, N, max(n.rep)))
ppc.out <- ppcOcc(out, 'chi-square', 1)
expect_type(ppc.out, "list")
expect_equal(dim(ppc.out$fit.y), c(n.post.samples, N))
expect_equal(dim(ppc.out$fit.y.rep), c(n.post.samples, N))
expect_equal(dim(ppc.out$fit.y.group.quants), c(5, N, J.fit))
expect_equal(dim(ppc.out$fit.y.rep.group.quants), c(5, N, J.fit))
ppc.out <- ppcOcc(out, 'freeman-tukey', 1)
expect_type(ppc.out, "list")
expect_equal(dim(ppc.out$fit.y), c(n.post.samples, N))
expect_equal(dim(ppc.out$fit.y.rep), c(n.post.samples, N))
expect_equal(dim(ppc.out$fit.y.group.quants), c(5, N, J.fit))
expect_equal(dim(ppc.out$fit.y.rep.group.quants), c(5, N, J.fit))
ppc.out <- ppcOcc(out, 'freeman-tukey', 2)
expect_type(ppc.out, "list")
expect_equal(dim(ppc.out$fit.y), c(n.post.samples, N))
expect_equal(dim(ppc.out$fit.y.rep), c(n.post.samples, N))
expect_equal(dim(ppc.out$fit.y.group.quants), c(5, N, max(n.rep)))
expect_equal(dim(ppc.out$fit.y.rep.group.quants), c(5, N, max(n.rep)))
})
# Occurrence coviarate only -----------------------------------------------
J.x <- 8
J.y <- 8
J <- J.x * J.y
n.rep<- sample(2:4, size = J, replace = TRUE)
N <- 6
# Community-level covariate effects
# Occurrence
beta.mean <- c(0.2, 0.5)
p.occ <- length(beta.mean)
tau.sq.beta <- c(0.6, 2.3)
# Detection
alpha.mean <- c(0)
tau.sq.alpha <- c(1)
p.det <- length(alpha.mean)
# Random effects
psi.RE <- list()
p.RE <- list()
# Draw species-level effects from community means.
beta <- matrix(NA, nrow = N, ncol = p.occ)
alpha <- matrix(NA, nrow = N, ncol = p.det)
for (i in 1:p.occ) {
beta[, i] <- rnorm(N, beta.mean[i], sqrt(tau.sq.beta[i]))
}
for (i in 1:p.det) {
alpha[, i] <- rnorm(N, alpha.mean[i], sqrt(tau.sq.alpha[i]))
}
alpha.true <- alpha
phi <- rep(3 / .7, N)
sigma.sq <- rep(2, N)
nu <- rep(2, N)
dat <- simMsOcc(J.x = J.x, J.y = J.y, n.rep = n.rep, N = N, beta = beta, alpha = alpha,
psi.RE = psi.RE, p.RE = p.RE, sp = TRUE, sigma.sq = sigma.sq,
phi = phi, nu = nu, cov.model = 'matern')
pred.indx <- sample(1:J, round(J * .25), replace = FALSE)
y <- dat$y[, -pred.indx, , drop = FALSE]
# Occupancy covariates
X <- dat$X[-pred.indx, , drop = FALSE]
coords <- as.matrix(dat$coords[-pred.indx, , drop = FALSE])
# Prediction covariates
X.0 <- dat$X[pred.indx, , drop = FALSE]
coords.0 <- as.matrix(dat$coords[pred.indx, , drop = FALSE])
# Detection covariates
X.p <- dat$X.p[-pred.indx, , , drop = FALSE]
occ.covs <- X
colnames(occ.covs) <- c('int', 'occ.cov.1')
data.list <- list(y = y, coords = coords, occ.covs = occ.covs)
# Priors
prior.list <- list(beta.comm.normal = list(mean = 0, var = 2.72),
alpha.comm.normal = list(mean = 0, var = 2.72),
tau.sq.beta.ig = list(a = 0.1, b = 0.1),
tau.sq.alpha.ig = list(a = 0.1, b = 0.1),
nu.unif = list(0.5, 2.5))
# Starting values
inits.list <- list(alpha.comm = 0,
beta.comm = 0,
beta = 0,
alpha = 0,
tau.sq.beta = 1,
tau.sq.alpha = 1,
z = apply(y, c(1, 2), max, na.rm = TRUE))
# Tuning
tuning.list <- list(phi = 1, nu = 0.25)
batch.length <- 25
n.batch <- 10
n.report <- 100
occ.formula <- ~ occ.cov.1
det.formula <- ~ 1
out <- spMsPGOcc(occ.formula = occ.formula,
det.formula = det.formula,
data = data.list,
inits = inits.list,
n.batch = n.batch,
batch.length = batch.length,
accept.rate = 0.43,
priors = prior.list,
cov.model = "matern",
tuning = tuning.list,
n.omp.threads = 1,
verbose = FALSE,
NNGP = TRUE,
n.neighbors = 5,
search.type = 'cb',
n.report = 10,
n.burn = 100,
n.thin = 2,
n.chains = 2,
k.fold = 2,
k.fold.threads = 1)
# To make sure it worked --------------
test_that("out is of class spMsPGOcc", {
expect_s3_class(out, "spMsPGOcc")
})
# Check cross-validation --------------
test_that("cross-validation works", {
expect_equal(length(out$k.fold.deviance), N)
expect_type(out$k.fold.deviance, "double")
expect_equal(sum(out$k.fold.deviance < 0), 0)
})
# Check random effects ----------------
test_that("random effects are correct", {
expect_equal(out$pRE, FALSE)
expect_equal(out$psiRE, FALSE)
})
# Check output data output is correct -
test_that("out$y == y", {
expect_equal(out$y, y)
})
# Check missing values ----------------
test_that("missing value error handling works", {
tmp.data <- data.list
tmp.data$occ.covs[3, ] <- NA
expect_error(spMsPGOcc(occ.formula = occ.formula,
det.formula = det.formula,
data = tmp.data,
n.batch = 40,
batch.length = batch.length,
cov.model = "exponential",
tuning = tuning.list,
NNGP = TRUE,
verbose = FALSE,
n.neighbors = 5,
search.type = 'cb',
n.report = 10,
n.burn = 100,
n.chains = 1))
# tmp.data <- data.list
# tmp.data$det.covs[[1]][1] <- NA
# expect_error(spMsPGOcc(occ.formula = occ.formula,
# det.formula = det.formula,
# data = tmp.data,
# n.batch = 40,
# batch.length = batch.length,
# cov.model = "exponential",
# tuning = tuning.list,
# NNGP = TRUE,
# verbose = FALSE,
# n.neighbors = 5,
# search.type = 'cb',
# n.report = 10,
# n.burn = 100,
# n.chains = 1))
# tmp.data <- data.list
# tmp.data$y[, 1, 1] <- NA
# out <- spMsPGOcc(occ.formula = occ.formula,
# det.formula = det.formula,
# data = tmp.data,
# n.batch = 40,
# batch.length = batch.length,
# cov.model = "exponential",
# tuning = tuning.list,
# NNGP = TRUE,
# verbose = FALSE,
# n.neighbors = 5,
# search.type = 'cb',
# n.report = 10,
# n.burn = 100,
# n.chains = 1)
# expect_s3_class(out, "spMsPGOcc")
})
# Check default values ----------------
test_that("default priors, inits, burn, thin work", {
out <- spMsPGOcc(occ.formula = occ.formula,
det.formula = det.formula,
data = data.list,
n.batch = n.batch,
batch.length = batch.length,
accept.rate = 0.43,
cov.model = "exponential",
tuning = tuning.list,
NNGP = TRUE,
verbose = FALSE,
n.neighbors = 5,
n.chains = 1)
expect_s3_class(out, "spMsPGOcc")
})
test_that("all correlation functions work", {
out <- spMsPGOcc(occ.formula = occ.formula,
det.formula = det.formula,
data = data.list,
inits = inits.list,
n.batch = n.batch,
batch.length = batch.length,
accept.rate = 0.43,
priors = prior.list,
cov.model = "gaussian",
tuning = list(phi = 0.3),
n.omp.threads = 1,
verbose = FALSE,
NNGP = TRUE,
n.neighbors = 5,
search.type = 'cb',
n.report = 10,
n.burn = 100,
n.thin = 1,
n.chains = 1)
expect_s3_class(out, "spMsPGOcc")
out <- spMsPGOcc(occ.formula = occ.formula,
det.formula = det.formula,
data = data.list,
inits = inits.list,
n.batch = n.batch,
batch.length = batch.length,
accept.rate = 0.43,
priors = prior.list,
cov.model = "spherical",
tuning = list(phi = 0.3),
n.omp.threads = 1,
verbose = FALSE,
NNGP = TRUE,
n.neighbors = 5,
search.type = 'cb',
n.report = 10,
n.burn = 100,
n.thin = 1,
n.chains = 1)
expect_s3_class(out, "spMsPGOcc")
out <- spMsPGOcc(occ.formula = occ.formula,
det.formula = det.formula,
data = data.list,
inits = inits.list,
n.batch = n.batch,
batch.length = batch.length,
accept.rate = 0.43,
priors = list(nu.unif = list(0.4, 3)),
cov.model = "matern",
tuning = list(phi = 0.3, nu = 0.2),
n.omp.threads = 1,
verbose = FALSE,
NNGP = TRUE,
n.neighbors = 5,
search.type = 'cb',
n.report = 10,
n.burn = 100,
n.thin = 1,
n.chains = 1)
expect_s3_class(out, "spMsPGOcc")
})
test_that("verbose prints to the screen", {
expect_output(spMsPGOcc(occ.formula = occ.formula,
det.formula = det.formula,
data = data.list,
inits = inits.list,
n.batch = n.batch,
batch.length = batch.length,
accept.rate = 0.43,
priors = prior.list,
cov.model = "exponential",
tuning = tuning.list,
n.omp.threads = 1,
verbose = TRUE,
NNGP = TRUE,
n.neighbors = 5,
search.type = 'cb',
n.report = 10,
n.burn = 100,
n.thin = 1,
n.chains = 1))
})
# Check waicOcc -----------------------
test_that("waicOCC works for spMsPGOcc", {
# as.vector gets rid of names
waic.out <- as.vector(waicOcc(out))
expect_equal(length(waic.out), 3)
expect_equal(waic.out[3], -2 * (waic.out[1] - waic.out[2]))
})
test_that("fitted works for spMsPGOcc", {
fitted.out <- fitted(out)
expect_equal(length(fitted.out), 2)
expect_equal(class(fitted.out$y.rep.samples), "array")
expect_equal(class(fitted.out$p.samples), "array")
expect_equal(dim(fitted.out$y.rep.samples), dim(fitted.out$p.samples))
})
test_that("predict works for spMsPGOcc", {
n.post.samples <- out$n.post * out$n.chains
pred.out <- predict(out, X.0, coords.0, verbose = FALSE)
expect_type(pred.out, "list")
expect_equal(dim(pred.out$psi.0.samples), c(n.post.samples, N, nrow(X.0)))
expect_equal(dim(pred.out$z.0.samples), c(n.post.samples, N, nrow(X.0)))
})
test_that("detection prediction works", {
J.str <- 100
X.p.0 <- matrix(1, nrow = J.str, ncol = p.det)
pred.out <- predict(out, X.p.0, type = 'detection')
expect_type(pred.out, 'list')
expect_equal(dim(pred.out$p.0.samples), c(out$n.post * out$n.chains, N, J.str))
})
test_that("posterior predictive checks work for msPGOcc", {
n.post.samples <- out$n.post * out$n.chains
J.fit <- nrow(X)
ppc.out <- ppcOcc(out, 'chi-square', 2)
expect_type(ppc.out, "list")
expect_equal(dim(ppc.out$fit.y), c(n.post.samples, N))
expect_equal(dim(ppc.out$fit.y.rep), c(n.post.samples, N))
expect_equal(dim(ppc.out$fit.y.group.quants), c(5, N, max(n.rep)))
expect_equal(dim(ppc.out$fit.y.rep.group.quants), c(5, N, max(n.rep)))
ppc.out <- ppcOcc(out, 'chi-square', 1)
expect_type(ppc.out, "list")
expect_equal(dim(ppc.out$fit.y), c(n.post.samples, N))
expect_equal(dim(ppc.out$fit.y.rep), c(n.post.samples, N))
expect_equal(dim(ppc.out$fit.y.group.quants), c(5, N, J.fit))
expect_equal(dim(ppc.out$fit.y.rep.group.quants), c(5, N, J.fit))
ppc.out <- ppcOcc(out, 'freeman-tukey', 1)
expect_type(ppc.out, "list")
expect_equal(dim(ppc.out$fit.y), c(n.post.samples, N))
expect_equal(dim(ppc.out$fit.y.rep), c(n.post.samples, N))
expect_equal(dim(ppc.out$fit.y.group.quants), c(5, N, J.fit))
expect_equal(dim(ppc.out$fit.y.rep.group.quants), c(5, N, J.fit))
ppc.out <- ppcOcc(out, 'freeman-tukey', 2)
expect_type(ppc.out, "list")
expect_equal(dim(ppc.out$fit.y), c(n.post.samples, N))
expect_equal(dim(ppc.out$fit.y.rep), c(n.post.samples, N))
expect_equal(dim(ppc.out$fit.y.group.quants), c(5, N, max(n.rep)))
expect_equal(dim(ppc.out$fit.y.rep.group.quants), c(5, N, max(n.rep)))
})
# Detection covariate only ------------------------------------------------
J.x <- 8
J.y <- 8
J <- J.x * J.y
n.rep<- sample(2:4, size = J, replace = TRUE)
N <- 6
# Community-level covariate effects
# Occurrence
beta.mean <- c(0.2)
p.occ <- length(beta.mean)
tau.sq.beta <- c(0.6)
# Detection
alpha.mean <- c(0, -0.5)
tau.sq.alpha <- c(1, 2.3)
p.det <- length(alpha.mean)
# Random effects
psi.RE <- list()
p.RE <- list()
# Draw species-level effects from community means.
beta <- matrix(NA, nrow = N, ncol = p.occ)
alpha <- matrix(NA, nrow = N, ncol = p.det)
for (i in 1:p.occ) {
beta[, i] <- rnorm(N, beta.mean[i], sqrt(tau.sq.beta[i]))
}
for (i in 1:p.det) {
alpha[, i] <- rnorm(N, alpha.mean[i], sqrt(tau.sq.alpha[i]))
}
alpha.true <- alpha
phi <- rep(3 / .7, N)
sigma.sq <- rep(2, N)
nu <- rep(2, N)
dat <- simMsOcc(J.x = J.x, J.y = J.y, n.rep = n.rep, N = N, beta = beta, alpha = alpha,
psi.RE = psi.RE, p.RE = p.RE, sp = TRUE, sigma.sq = sigma.sq,
phi = phi, nu = nu, cov.model = 'matern')
pred.indx <- sample(1:J, round(J * .25), replace = FALSE)
y <- dat$y[, -pred.indx, , drop = FALSE]
# Occupancy covariates
X <- dat$X[-pred.indx, , drop = FALSE]
coords <- as.matrix(dat$coords[-pred.indx, , drop = FALSE])
# Prediction covariates
X.0 <- dat$X[pred.indx, , drop = FALSE]
coords.0 <- as.matrix(dat$coords[pred.indx, , drop = FALSE])
# Detection covariates
X.p <- dat$X.p[-pred.indx, , , drop = FALSE]
occ.covs <- X
colnames(occ.covs) <- c('int')
det.covs <- list(det.cov.1 = X.p[, , 2])
data.list <- list(y = y, coords = coords, occ.covs = occ.covs, det.covs = det.covs)
# Priors
prior.list <- list(beta.comm.normal = list(mean = 0, var = 2.72),
alpha.comm.normal = list(mean = 0, var = 2.72),
tau.sq.beta.ig = list(a = 0.1, b = 0.1),
tau.sq.alpha.ig = list(a = 0.1, b = 0.1),
nu.unif = list(0.5, 2.5))
# Starting values
inits.list <- list(alpha.comm = 0,
beta.comm = 0,
beta = 0,
alpha = 0,
tau.sq.beta = 1,
tau.sq.alpha = 1,
z = apply(y, c(1, 2), max, na.rm = TRUE))
# Tuning
tuning.list <- list(phi = 1, nu = 0.25)
batch.length <- 25
n.batch <- 10
n.report <- 100
occ.formula <- ~ 1
det.formula <- ~ det.cov.1
out <- spMsPGOcc(occ.formula = occ.formula,
det.formula = det.formula,
data = data.list,
inits = inits.list,
n.batch = n.batch,
batch.length = batch.length,
accept.rate = 0.43,
priors = prior.list,
cov.model = "matern",
tuning = tuning.list,
n.omp.threads = 1,
verbose = FALSE,
NNGP = TRUE,
n.neighbors = 5,
search.type = 'cb',
n.report = 10,
n.burn = 100,
n.thin = 2,
n.chains = 2,
k.fold = 2,
k.fold.threads = 1)
# To make sure it worked --------------
test_that("out is of class spMsPGOcc", {
expect_s3_class(out, "spMsPGOcc")
})
# Check cross-validation --------------
test_that("cross-validation works", {
expect_equal(length(out$k.fold.deviance), N)
expect_type(out$k.fold.deviance, "double")
expect_equal(sum(out$k.fold.deviance < 0), 0)
})
# Check random effects ----------------
test_that("random effects are correct", {
expect_equal(out$pRE, FALSE)
expect_equal(out$psiRE, FALSE)
})
# Check output data output is correct -
test_that("out$y == y", {
expect_equal(out$y, y)
})
# Check missing values ----------------
test_that("missing value error handling works", {
# tmp.data <- data.list
# tmp.data$occ.covs[3, ] <- NA
# expect_error(spMsPGOcc(occ.formula = occ.formula,
# det.formula = det.formula,
# data = tmp.data,
# n.batch = 40,
# batch.length = batch.length,
# cov.model = "exponential",
# tuning = tuning.list,
# NNGP = TRUE,
# verbose = FALSE,
# n.neighbors = 5,
# search.type = 'cb',
# n.report = 10,
# n.burn = 100,
# n.chains = 1))
tmp.data <- data.list
tmp.data$y[1, 1, 1] <- 1
tmp.data$det.covs[[1]][1] <- NA
expect_error(spMsPGOcc(occ.formula = occ.formula,
det.formula = det.formula,
data = tmp.data,
n.batch = 40,
batch.length = batch.length,
cov.model = "exponential",
tuning = tuning.list,
NNGP = TRUE,
verbose = FALSE,
n.neighbors = 5,
search.type = 'cb',
n.report = 10,
n.burn = 100,
n.chains = 1))
tmp.data <- data.list
tmp.data$y[, 1, 1] <- NA
out <- spMsPGOcc(occ.formula = occ.formula,
det.formula = det.formula,
data = tmp.data,
n.batch = 40,
batch.length = batch.length,
cov.model = "exponential",
tuning = tuning.list,
NNGP = TRUE,
verbose = FALSE,
n.neighbors = 5,
search.type = 'cb',
n.report = 10,
n.burn = 100,
n.chains = 1)
expect_s3_class(out, "spMsPGOcc")
})
# Check default values ----------------
test_that("default priors, inits, burn, thin work", {
out <- spMsPGOcc(occ.formula = occ.formula,
det.formula = det.formula,
data = data.list,
n.batch = n.batch,
batch.length = batch.length,
accept.rate = 0.43,
cov.model = "exponential",
tuning = tuning.list,
NNGP = TRUE,
verbose = FALSE,
n.neighbors = 5,
n.chains = 1)
expect_s3_class(out, "spMsPGOcc")
})
test_that("all correlation functions work", {
out <- spMsPGOcc(occ.formula = occ.formula,
det.formula = det.formula,
data = data.list,
inits = inits.list,
n.batch = n.batch,
batch.length = batch.length,
accept.rate = 0.43,
priors = prior.list,
cov.model = "gaussian",
tuning = list(phi = 0.3),
n.omp.threads = 1,
verbose = FALSE,
NNGP = TRUE,
n.neighbors = 5,
search.type = 'cb',
n.report = 10,
n.burn = 100,
n.thin = 1,
n.chains = 1)
expect_s3_class(out, "spMsPGOcc")
out <- spMsPGOcc(occ.formula = occ.formula,
det.formula = det.formula,
data = data.list,
inits = inits.list,
n.batch = n.batch,
batch.length = batch.length,
accept.rate = 0.43,
priors = prior.list,
cov.model = "spherical",
tuning = list(phi = 0.3),
n.omp.threads = 1,
verbose = FALSE,
NNGP = TRUE,
n.neighbors = 5,
search.type = 'cb',
n.report = 10,
n.burn = 100,
n.thin = 1,
n.chains = 1)
expect_s3_class(out, "spMsPGOcc")
out <- spMsPGOcc(occ.formula = occ.formula,
det.formula = det.formula,
data = data.list,
inits = inits.list,
n.batch = n.batch,
batch.length = batch.length,
accept.rate = 0.43,
priors = list(nu.unif = list(0.4, 3)),
cov.model = "matern",
tuning = list(phi = 0.3, nu = 0.2),
n.omp.threads = 1,
verbose = FALSE,
NNGP = TRUE,
n.neighbors = 5,
search.type = 'cb',
n.report = 10,
n.burn = 100,
n.thin = 1,
n.chains = 1)
expect_s3_class(out, "spMsPGOcc")
})
test_that("verbose prints to the screen", {
expect_output(spMsPGOcc(occ.formula = occ.formula,
det.formula = det.formula,
data = data.list,
inits = inits.list,
n.batch = n.batch,
batch.length = batch.length,
accept.rate = 0.43,
priors = prior.list,
cov.model = "exponential",
tuning = tuning.list,
n.omp.threads = 1,
verbose = TRUE,
NNGP = TRUE,
n.neighbors = 5,
search.type = 'cb',
n.report = 10,
n.burn = 100,
n.thin = 1,
n.chains = 1))
})
# Check waicOcc -----------------------
test_that("waicOCC works for spMsPGOcc", {
# as.vector gets rid of names
waic.out <- as.vector(waicOcc(out))
expect_equal(length(waic.out), 3)
expect_equal(waic.out[3], -2 * (waic.out[1] - waic.out[2]))
})
test_that("fitted works for spMsPGOcc", {
fitted.out <- fitted(out)
expect_equal(length(fitted.out), 2)
expect_equal(class(fitted.out$y.rep.samples), "array")
expect_equal(class(fitted.out$p.samples), "array")
expect_equal(dim(fitted.out$y.rep.samples), dim(fitted.out$p.samples))
})
test_that("predict works for spMsPGOcc", {
n.post.samples <- out$n.post * out$n.chains
pred.out <- predict(out, X.0, coords.0, verbose = FALSE)
expect_type(pred.out, "list")
expect_equal(dim(pred.out$psi.0.samples), c(n.post.samples, N, nrow(X.0)))
expect_equal(dim(pred.out$z.0.samples), c(n.post.samples, N, nrow(X.0)))
})
test_that("detection prediction works", {
X.p.0 <- dat$X.p[, 1, ]
pred.out <- predict(out, X.p.0, type = 'detection')
expect_type(pred.out, 'list')
expect_equal(dim(pred.out$p.0.samples), c(out$n.post * out$n.chains, N, J))
})
test_that("posterior predictive checks work for msPGOcc", {
n.post.samples <- out$n.post * out$n.chains
J.fit <- nrow(X)
ppc.out <- ppcOcc(out, 'chi-square', 2)
expect_type(ppc.out, "list")
expect_equal(dim(ppc.out$fit.y), c(n.post.samples, N))
expect_equal(dim(ppc.out$fit.y.rep), c(n.post.samples, N))
expect_equal(dim(ppc.out$fit.y.group.quants), c(5, N, max(n.rep)))
expect_equal(dim(ppc.out$fit.y.rep.group.quants), c(5, N, max(n.rep)))
ppc.out <- ppcOcc(out, 'chi-square', 1)
expect_type(ppc.out, "list")
expect_equal(dim(ppc.out$fit.y), c(n.post.samples, N))
expect_equal(dim(ppc.out$fit.y.rep), c(n.post.samples, N))
expect_equal(dim(ppc.out$fit.y.group.quants), c(5, N, J.fit))
expect_equal(dim(ppc.out$fit.y.rep.group.quants), c(5, N, J.fit))
ppc.out <- ppcOcc(out, 'freeman-tukey', 1)
expect_type(ppc.out, "list")
expect_equal(dim(ppc.out$fit.y), c(n.post.samples, N))
expect_equal(dim(ppc.out$fit.y.rep), c(n.post.samples, N))
expect_equal(dim(ppc.out$fit.y.group.quants), c(5, N, J.fit))
expect_equal(dim(ppc.out$fit.y.rep.group.quants), c(5, N, J.fit))
ppc.out <- ppcOcc(out, 'freeman-tukey', 2)
expect_type(ppc.out, "list")
expect_equal(dim(ppc.out$fit.y), c(n.post.samples, N))
expect_equal(dim(ppc.out$fit.y.rep), c(n.post.samples, N))
expect_equal(dim(ppc.out$fit.y.group.quants), c(5, N, max(n.rep)))
expect_equal(dim(ppc.out$fit.y.rep.group.quants), c(5, N, max(n.rep)))
})
# Covariates on both ------------------------------------------------------
J.x <- 8
J.y <- 8
J <- J.x * J.y
n.rep<- sample(2:4, size = J, replace = TRUE)
N <- 6
# Community-level covariate effects
# Occurrence
beta.mean <- c(0.2, 0.5, 1.2)
p.occ <- length(beta.mean)
tau.sq.beta <- c(0.6, 1.5, 2.3)
# Detection
alpha.mean <- c(0, -0.5)
tau.sq.alpha <- c(1, 2.3)
p.det <- length(alpha.mean)
# Random effects
psi.RE <- list()
p.RE <- list()
# Draw species-level effects from community means.
beta <- matrix(NA, nrow = N, ncol = p.occ)
alpha <- matrix(NA, nrow = N, ncol = p.det)
for (i in 1:p.occ) {
beta[, i] <- rnorm(N, beta.mean[i], sqrt(tau.sq.beta[i]))
}
for (i in 1:p.det) {
alpha[, i] <- rnorm(N, alpha.mean[i], sqrt(tau.sq.alpha[i]))
}
alpha.true <- alpha
phi <- rep(3 / .7, N)
sigma.sq <- rep(2, N)
nu <- rep(2, N)
dat <- simMsOcc(J.x = J.x, J.y = J.y, n.rep = n.rep, N = N, beta = beta, alpha = alpha,
psi.RE = psi.RE, p.RE = p.RE, sp = TRUE, sigma.sq = sigma.sq,
phi = phi, nu = nu, cov.model = 'matern')
pred.indx <- sample(1:J, round(J * .25), replace = FALSE)
y <- dat$y[, -pred.indx, , drop = FALSE]
# Occupancy covariates
X <- dat$X[-pred.indx, , drop = FALSE]
coords <- as.matrix(dat$coords[-pred.indx, , drop = FALSE])
# Prediction covariates
X.0 <- dat$X[pred.indx, , drop = FALSE]
coords.0 <- as.matrix(dat$coords[pred.indx, , drop = FALSE])
# Detection covariates
X.p <- dat$X.p[-pred.indx, , , drop = FALSE]
occ.covs <- X
colnames(occ.covs) <- c('int', 'occ.cov.1', 'occ.cov.2')
det.covs <- list(det.cov.1 = X.p[, , 2])
data.list <- list(y = y, coords = coords, occ.covs = occ.covs, det.covs = det.covs)
# Priors
prior.list <- list(beta.comm.normal = list(mean = 0, var = 2.72),
alpha.comm.normal = list(mean = 0, var = 2.72),
tau.sq.beta.ig = list(a = 0.1, b = 0.1),
tau.sq.alpha.ig = list(a = 0.1, b = 0.1),
nu.unif = list(0.5, 2.5))
# Starting values
inits.list <- list(alpha.comm = 0,
beta.comm = 0,
beta = 0,
alpha = 0,
tau.sq.beta = 1,
tau.sq.alpha = 1,
z = apply(y, c(1, 2), max, na.rm = TRUE))
# Tuning
tuning.list <- list(phi = 1, nu = 0.25)
batch.length <- 25
n.batch <- 10
n.report <- 100
occ.formula <- ~ occ.cov.1 + occ.cov.2
det.formula <- ~ det.cov.1
out <- spMsPGOcc(occ.formula = occ.formula,
det.formula = det.formula,
data = data.list,
inits = inits.list,
n.batch = n.batch,
batch.length = batch.length,
accept.rate = 0.43,
priors = prior.list,
cov.model = "matern",
tuning = tuning.list,
n.omp.threads = 1,
verbose = FALSE,
NNGP = TRUE,
n.neighbors = 5,
search.type = 'cb',
n.report = 10,
n.burn = 100,
n.thin = 2,
n.chains = 2,
k.fold = 2,
k.fold.threads = 1)
# To make sure it worked --------------
test_that("out is of class spMsPGOcc", {
expect_s3_class(out, "spMsPGOcc")
})
# Check cross-validation --------------
test_that("cross-validation works", {
expect_equal(length(out$k.fold.deviance), N)
expect_type(out$k.fold.deviance, "double")
expect_equal(sum(out$k.fold.deviance < 0), 0)
})
# Check random effects ----------------
test_that("random effects are correct", {
expect_equal(out$pRE, FALSE)
expect_equal(out$psiRE, FALSE)
})
# Check output data output is correct -
test_that("out$y == y", {
expect_equal(out$y, y)
})
# Check missing values ----------------
test_that("missing value error handling works", {
tmp.data <- data.list
tmp.data$occ.covs[3, ] <- NA
expect_error(spMsPGOcc(occ.formula = occ.formula,
det.formula = det.formula,
data = tmp.data,
n.batch = 40,
batch.length = batch.length,
cov.model = "exponential",
tuning = tuning.list,
NNGP = TRUE,
verbose = FALSE,
n.neighbors = 5,
search.type = 'cb',
n.report = 10,
n.burn = 100,
n.chains = 1))
tmp.data <- data.list
tmp.data$y[1, 1, 1] <- 1
tmp.data$det.covs[[1]][1] <- NA
expect_error(spMsPGOcc(occ.formula = occ.formula,
det.formula = det.formula,
data = tmp.data,
n.batch = 40,
batch.length = batch.length,
cov.model = "exponential",
tuning = tuning.list,
NNGP = TRUE,
verbose = FALSE,
n.neighbors = 5,
search.type = 'cb',
n.report = 10,
n.burn = 100,
n.chains = 1))
tmp.data <- data.list
tmp.data$y[, 1, 1] <- NA
out <- spMsPGOcc(occ.formula = occ.formula,
det.formula = det.formula,
data = tmp.data,
n.batch = 40,
batch.length = batch.length,
cov.model = "exponential",
tuning = tuning.list,
NNGP = TRUE,
verbose = FALSE,
n.neighbors = 5,
search.type = 'cb',
n.report = 10,
n.burn = 100,
n.chains = 1)
expect_s3_class(out, "spMsPGOcc")
})
# Check default values ----------------
test_that("default priors, inits, burn, thin work", {
out <- spMsPGOcc(occ.formula = occ.formula,
det.formula = det.formula,
data = data.list,
n.batch = n.batch,
batch.length = batch.length,
accept.rate = 0.43,
cov.model = "exponential",
tuning = tuning.list,
NNGP = TRUE,
verbose = FALSE,
n.neighbors = 5,
n.chains = 1)
expect_s3_class(out, "spMsPGOcc")
})
test_that("all correlation functions work", {
out <- spMsPGOcc(occ.formula = occ.formula,
det.formula = det.formula,
data = data.list,
inits = inits.list,
n.batch = n.batch,
batch.length = batch.length,
accept.rate = 0.43,
priors = prior.list,
cov.model = "gaussian",
tuning = list(phi = 0.3),
n.omp.threads = 1,
verbose = FALSE,
NNGP = TRUE,
n.neighbors = 5,
search.type = 'cb',
n.report = 10,
n.burn = 100,
n.thin = 1,
n.chains = 1)
expect_s3_class(out, "spMsPGOcc")
out <- spMsPGOcc(occ.formula = occ.formula,
det.formula = det.formula,
data = data.list,
inits = inits.list,
n.batch = n.batch,
batch.length = batch.length,
accept.rate = 0.43,
priors = prior.list,
cov.model = "spherical",
tuning = list(phi = 0.3),
n.omp.threads = 1,
verbose = FALSE,
NNGP = TRUE,
n.neighbors = 5,
search.type = 'cb',
n.report = 10,
n.burn = 100,
n.thin = 1,
n.chains = 1)
expect_s3_class(out, "spMsPGOcc")
out <- spMsPGOcc(occ.formula = occ.formula,
det.formula = det.formula,
data = data.list,
inits = inits.list,
n.batch = n.batch,
batch.length = batch.length,
accept.rate = 0.43,
priors = list(nu.unif = list(0.4, 3)),
cov.model = "matern",
tuning = list(phi = 0.3, nu = 0.2),
n.omp.threads = 1,
verbose = FALSE,
NNGP = TRUE,
n.neighbors = 5,
search.type = 'cb',
n.report = 10,
n.burn = 100,
n.thin = 1,
n.chains = 1)
expect_s3_class(out, "spMsPGOcc")
})
test_that("verbose prints to the screen", {
expect_output(spMsPGOcc(occ.formula = occ.formula,
det.formula = det.formula,
data = data.list,
inits = inits.list,
n.batch = n.batch,
batch.length = batch.length,
accept.rate = 0.43,
priors = prior.list,
cov.model = "exponential",
tuning = tuning.list,
n.omp.threads = 1,
verbose = TRUE,
NNGP = TRUE,
n.neighbors = 5,
search.type = 'cb',
n.report = 10,
n.burn = 100,
n.thin = 1,
n.chains = 1))
})
# Check waicOcc -----------------------
test_that("waicOCC works for spMsPGOcc", {
# as.vector gets rid of names
waic.out <- as.vector(waicOcc(out))
expect_equal(length(waic.out), 3)
expect_equal(waic.out[3], -2 * (waic.out[1] - waic.out[2]))
})
test_that("fitted works for spMsPGOcc", {
fitted.out <- fitted(out)
expect_equal(length(fitted.out), 2)
expect_equal(class(fitted.out$y.rep.samples), "array")
expect_equal(class(fitted.out$p.samples), "array")
expect_equal(dim(fitted.out$y.rep.samples), dim(fitted.out$p.samples))
})
test_that("predict works for spMsPGOcc", {
n.post.samples <- out$n.post * out$n.chains
pred.out <- predict(out, X.0, coords.0, verbose = FALSE)
expect_type(pred.out, "list")
expect_equal(dim(pred.out$psi.0.samples), c(n.post.samples, N, nrow(X.0)))
expect_equal(dim(pred.out$z.0.samples), c(n.post.samples, N, nrow(X.0)))
})
test_that("detection prediction works", {
X.p.0 <- dat$X.p[, 1, ]
pred.out <- predict(out, X.p.0, type = 'detection')
expect_type(pred.out, 'list')
expect_equal(dim(pred.out$p.0.samples), c(out$n.post * out$n.chains, N, J))
})
test_that("posterior predictive checks work for msPGOcc", {
n.post.samples <- out$n.post * out$n.chains
J.fit <- nrow(X)
ppc.out <- ppcOcc(out, 'chi-square', 2)
expect_type(ppc.out, "list")
expect_equal(dim(ppc.out$fit.y), c(n.post.samples, N))
expect_equal(dim(ppc.out$fit.y.rep), c(n.post.samples, N))
expect_equal(dim(ppc.out$fit.y.group.quants), c(5, N, max(n.rep)))
expect_equal(dim(ppc.out$fit.y.rep.group.quants), c(5, N, max(n.rep)))
ppc.out <- ppcOcc(out, 'chi-square', 1)
expect_type(ppc.out, "list")
expect_equal(dim(ppc.out$fit.y), c(n.post.samples, N))
expect_equal(dim(ppc.out$fit.y.rep), c(n.post.samples, N))
expect_equal(dim(ppc.out$fit.y.group.quants), c(5, N, J.fit))
expect_equal(dim(ppc.out$fit.y.rep.group.quants), c(5, N, J.fit))
ppc.out <- ppcOcc(out, 'freeman-tukey', 1)
expect_type(ppc.out, "list")
expect_equal(dim(ppc.out$fit.y), c(n.post.samples, N))
expect_equal(dim(ppc.out$fit.y.rep), c(n.post.samples, N))
expect_equal(dim(ppc.out$fit.y.group.quants), c(5, N, J.fit))
expect_equal(dim(ppc.out$fit.y.rep.group.quants), c(5, N, J.fit))
ppc.out <- ppcOcc(out, 'freeman-tukey', 2)
expect_type(ppc.out, "list")
expect_equal(dim(ppc.out$fit.y), c(n.post.samples, N))
expect_equal(dim(ppc.out$fit.y.rep), c(n.post.samples, N))
expect_equal(dim(ppc.out$fit.y.group.quants), c(5, N, max(n.rep)))
expect_equal(dim(ppc.out$fit.y.rep.group.quants), c(5, N, max(n.rep)))
})
# Interactions on both ----------------------------------------------------
J.x <- 8
J.y <- 8
J <- J.x * J.y
n.rep<- sample(2:4, size = J, replace = TRUE)
N <- 6
# Community-level covariate effects
# Occurrence
beta.mean <- c(0.2, 0.5, 1.2)
p.occ <- length(beta.mean)
tau.sq.beta <- c(0.6, 1.5, 2.3)
# Detection
alpha.mean <- c(0, -0.5, 1.2)
tau.sq.alpha <- c(1, 2.3, 1.5)
p.det <- length(alpha.mean)
# Random effects
psi.RE <- list()
p.RE <- list()
# Draw species-level effects from community means.
beta <- matrix(NA, nrow = N, ncol = p.occ)
alpha <- matrix(NA, nrow = N, ncol = p.det)
for (i in 1:p.occ) {
beta[, i] <- rnorm(N, beta.mean[i], sqrt(tau.sq.beta[i]))
}
for (i in 1:p.det) {
alpha[, i] <- rnorm(N, alpha.mean[i], sqrt(tau.sq.alpha[i]))
}
alpha.true <- alpha
phi <- rep(3 / .7, N)
sigma.sq <- rep(2, N)
nu <- rep(2, N)
dat <- simMsOcc(J.x = J.x, J.y = J.y, n.rep = n.rep, N = N, beta = beta, alpha = alpha,
psi.RE = psi.RE, p.RE = p.RE, sp = TRUE, sigma.sq = sigma.sq,
phi = phi, nu = nu, cov.model = 'matern')
pred.indx <- sample(1:J, round(J * .25), replace = FALSE)
y <- dat$y[, -pred.indx, , drop = FALSE]
# Occupancy covariates
X <- dat$X[-pred.indx, , drop = FALSE]
coords <- as.matrix(dat$coords[-pred.indx, , drop = FALSE])
# Prediction covariates
X.0 <- dat$X[pred.indx, , drop = FALSE]
coords.0 <- as.matrix(dat$coords[pred.indx, , drop = FALSE])
# Detection covariates
X.p <- dat$X.p[-pred.indx, , , drop = FALSE]
occ.covs <- X
colnames(occ.covs) <- c('int', 'occ.cov.1', 'occ.cov.2')
det.covs <- list(det.cov.1 = X.p[, , 2],
det.cov.2 = X.p[, , 3])
data.list <- list(y = y, coords = coords, occ.covs = occ.covs, det.covs = det.covs)
# Priors
prior.list <- list(beta.comm.normal = list(mean = 0, var = 2.72),
alpha.comm.normal = list(mean = 0, var = 2.72),
tau.sq.beta.ig = list(a = 0.1, b = 0.1),
tau.sq.alpha.ig = list(a = 0.1, b = 0.1),
nu.unif = list(0.5, 2.5))
# Starting values
inits.list <- list(alpha.comm = 0,
beta.comm = 0,
beta = 0,
alpha = 0,
tau.sq.beta = 1,
tau.sq.alpha = 1,
z = apply(y, c(1, 2), max, na.rm = TRUE))
# Tuning
tuning.list <- list(phi = 1, nu = 0.25)
batch.length <- 25
n.batch <- 10
n.report <- 100
occ.formula <- ~ occ.cov.1 * occ.cov.2
det.formula <- ~ det.cov.1 * det.cov.2
out <- spMsPGOcc(occ.formula = occ.formula,
det.formula = det.formula,
data = data.list,
inits = inits.list,
n.batch = n.batch,
batch.length = batch.length,
accept.rate = 0.43,
priors = prior.list,
cov.model = "matern",
tuning = tuning.list,
n.omp.threads = 1,
verbose = FALSE,
NNGP = TRUE,
n.neighbors = 5,
search.type = 'cb',
n.report = 10,
n.burn = 100,
n.thin = 2,
n.chains = 2,
k.fold = 2,
k.fold.threads = 1)
# To make sure it worked --------------
test_that("out is of class spMsPGOcc", {
expect_s3_class(out, "spMsPGOcc")
})
# Check cross-validation --------------
test_that("cross-validation works", {
expect_equal(length(out$k.fold.deviance), N)
expect_type(out$k.fold.deviance, "double")
expect_equal(sum(out$k.fold.deviance < 0), 0)
})
# Check random effects ----------------
test_that("random effects are correct", {
expect_equal(out$pRE, FALSE)
expect_equal(out$psiRE, FALSE)
})
# Check output data output is correct -
test_that("out$y == y", {
expect_equal(out$y, y)
})
# Check missing values ----------------
test_that("missing value error handling works", {
tmp.data <- data.list
tmp.data$occ.covs[3, ] <- NA
expect_error(spMsPGOcc(occ.formula = occ.formula,
det.formula = det.formula,
data = tmp.data,
n.batch = 40,
batch.length = batch.length,
cov.model = "exponential",
tuning = tuning.list,
NNGP = TRUE,
verbose = FALSE,
n.neighbors = 5,
search.type = 'cb',
n.report = 10,
n.burn = 100,
n.chains = 1))
tmp.data <- data.list
tmp.data$y[1, 1, 1] <- 1
tmp.data$det.covs[[1]][1] <- NA
expect_error(spMsPGOcc(occ.formula = occ.formula,
det.formula = det.formula,
data = tmp.data,
n.batch = 40,
batch.length = batch.length,
cov.model = "exponential",
tuning = tuning.list,
NNGP = TRUE,
verbose = FALSE,
n.neighbors = 5,
search.type = 'cb',
n.report = 10,
n.burn = 100,
n.chains = 1))
tmp.data <- data.list
tmp.data$y[, 1, 1] <- NA
out <- spMsPGOcc(occ.formula = occ.formula,
det.formula = det.formula,
data = tmp.data,
n.batch = 40,
batch.length = batch.length,
cov.model = "exponential",
tuning = tuning.list,
NNGP = TRUE,
verbose = FALSE,
n.neighbors = 5,
search.type = 'cb',
n.report = 10,
n.burn = 100,
n.chains = 1)
expect_s3_class(out, "spMsPGOcc")
})
# Check default values ----------------
test_that("default priors, inits, burn, thin work", {
out <- spMsPGOcc(occ.formula = occ.formula,
det.formula = det.formula,
data = data.list,
n.batch = n.batch,
batch.length = batch.length,
accept.rate = 0.43,
cov.model = "exponential",
tuning = tuning.list,
NNGP = TRUE,
verbose = FALSE,
n.neighbors = 5,
n.chains = 1)
expect_s3_class(out, "spMsPGOcc")
})
test_that("all correlation functions work", {
out <- spMsPGOcc(occ.formula = occ.formula,
det.formula = det.formula,
data = data.list,
inits = inits.list,
n.batch = n.batch,
batch.length = batch.length,
accept.rate = 0.43,
priors = prior.list,
cov.model = "gaussian",
tuning = list(phi = 0.3),
n.omp.threads = 1,
verbose = FALSE,
NNGP = TRUE,
n.neighbors = 5,
search.type = 'cb',
n.report = 10,
n.burn = 100,
n.thin = 1,
n.chains = 1)
expect_s3_class(out, "spMsPGOcc")
out <- spMsPGOcc(occ.formula = occ.formula,
det.formula = det.formula,
data = data.list,
inits = inits.list,
n.batch = n.batch,
batch.length = batch.length,
accept.rate = 0.43,
priors = prior.list,
cov.model = "spherical",
tuning = list(phi = 0.3),
n.omp.threads = 1,
verbose = FALSE,
NNGP = TRUE,
n.neighbors = 5,
search.type = 'cb',
n.report = 10,
n.burn = 100,
n.thin = 1,
n.chains = 1)
expect_s3_class(out, "spMsPGOcc")
out <- spMsPGOcc(occ.formula = occ.formula,
det.formula = det.formula,
data = data.list,
inits = inits.list,
n.batch = n.batch,
batch.length = batch.length,
accept.rate = 0.43,
priors = list(nu.unif = list(0.4, 3)),
cov.model = "matern",
tuning = list(phi = 0.3, nu = 0.2),
n.omp.threads = 1,
verbose = FALSE,
NNGP = TRUE,
n.neighbors = 5,
search.type = 'cb',
n.report = 10,
n.burn = 100,
n.thin = 1,
n.chains = 1)
expect_s3_class(out, "spMsPGOcc")
})
test_that("verbose prints to the screen", {
expect_output(spMsPGOcc(occ.formula = occ.formula,
det.formula = det.formula,
data = data.list,
inits = inits.list,
n.batch = n.batch,
batch.length = batch.length,
accept.rate = 0.43,
priors = prior.list,
cov.model = "exponential",
tuning = tuning.list,
n.omp.threads = 1,
verbose = TRUE,
NNGP = TRUE,
n.neighbors = 5,
search.type = 'cb',
n.report = 10,
n.burn = 100,
n.thin = 1,
n.chains = 1))
})
# Check waicOcc -----------------------
test_that("waicOCC works for spMsPGOcc", {
# as.vector gets rid of names
waic.out <- as.vector(waicOcc(out))
expect_equal(length(waic.out), 3)
expect_equal(waic.out[3], -2 * (waic.out[1] - waic.out[2]))
})
test_that("fitted works for spMsPGOcc", {
fitted.out <- fitted(out)
expect_equal(length(fitted.out), 2)
expect_equal(class(fitted.out$y.rep.samples), "array")
expect_equal(class(fitted.out$p.samples), "array")
expect_equal(dim(fitted.out$y.rep.samples), dim(fitted.out$p.samples))
})
test_that("predict works for spMsPGOcc", {
n.post.samples <- out$n.post * out$n.chains
X.0 <- cbind(X.0, X.0[, 2] * X.0[, 3])
colnames(X.0) <- c('int', 'occ.cov.1', 'occ.cov.2', 'occ.cov.1:occ.cov.2')
pred.out <- predict(out, X.0, coords.0, verbose = FALSE)
expect_type(pred.out, "list")
expect_equal(dim(pred.out$psi.0.samples), c(n.post.samples, N, nrow(X.0)))
expect_equal(dim(pred.out$z.0.samples), c(n.post.samples, N, nrow(X.0)))
})
test_that("detection prediction works", {
X.p.0 <- dat$X.p[, 1, ]
X.p.0 <- cbind(X.p.0, X.p.0[, 2] * X.p.0[, 3])
pred.out <- predict(out, X.p.0, type = 'detection')
expect_type(pred.out, 'list')
expect_equal(dim(pred.out$p.0.samples), c(out$n.post * out$n.chains, N, J))
})
test_that("posterior predictive checks work for msPGOcc", {
n.post.samples <- out$n.post * out$n.chains
J.fit <- nrow(X)
ppc.out <- ppcOcc(out, 'chi-square', 2)
expect_type(ppc.out, "list")
expect_equal(dim(ppc.out$fit.y), c(n.post.samples, N))
expect_equal(dim(ppc.out$fit.y.rep), c(n.post.samples, N))
expect_equal(dim(ppc.out$fit.y.group.quants), c(5, N, max(n.rep)))
expect_equal(dim(ppc.out$fit.y.rep.group.quants), c(5, N, max(n.rep)))
ppc.out <- ppcOcc(out, 'chi-square', 1)
expect_type(ppc.out, "list")
expect_equal(dim(ppc.out$fit.y), c(n.post.samples, N))
expect_equal(dim(ppc.out$fit.y.rep), c(n.post.samples, N))
expect_equal(dim(ppc.out$fit.y.group.quants), c(5, N, J.fit))
expect_equal(dim(ppc.out$fit.y.rep.group.quants), c(5, N, J.fit))
ppc.out <- ppcOcc(out, 'freeman-tukey', 1)
expect_type(ppc.out, "list")
expect_equal(dim(ppc.out$fit.y), c(n.post.samples, N))
expect_equal(dim(ppc.out$fit.y.rep), c(n.post.samples, N))
expect_equal(dim(ppc.out$fit.y.group.quants), c(5, N, J.fit))
expect_equal(dim(ppc.out$fit.y.rep.group.quants), c(5, N, J.fit))
ppc.out <- ppcOcc(out, 'freeman-tukey', 2)
expect_type(ppc.out, "list")
expect_equal(dim(ppc.out$fit.y), c(n.post.samples, N))
expect_equal(dim(ppc.out$fit.y.rep), c(n.post.samples, N))
expect_equal(dim(ppc.out$fit.y.group.quants), c(5, N, max(n.rep)))
expect_equal(dim(ppc.out$fit.y.rep.group.quants), c(5, N, max(n.rep)))
})
# Site covariate on detection ---------------------------------------------
set.seed(400)
J.x <- 8
J.y <- 8
J <- J.x * J.y
n.rep<- sample(2:4, size = J, replace = TRUE)
N <- 6
# Community-level covariate effects
# Occurrence
beta.mean <- c(0.2, 0.5, 1.2)
p.occ <- length(beta.mean)
tau.sq.beta <- c(0.6, 1.5, 2.3)
# Detection
alpha.mean <- c(0, -0.5)
tau.sq.alpha <- c(1, 2.3)
p.det <- length(alpha.mean)
# Random effects
psi.RE <- list()
p.RE <- list()
# Draw species-level effects from community means.
beta <- matrix(NA, nrow = N, ncol = p.occ)
alpha <- matrix(NA, nrow = N, ncol = p.det)
for (i in 1:p.occ) {
beta[, i] <- rnorm(N, beta.mean[i], sqrt(tau.sq.beta[i]))
}
for (i in 1:p.det) {
alpha[, i] <- rnorm(N, alpha.mean[i], sqrt(tau.sq.alpha[i]))
}
alpha.true <- alpha
phi <- rep(3 / .7, N)
sigma.sq <- rep(2, N)
nu <- rep(2, N)
dat <- simMsOcc(J.x = J.x, J.y = J.y, n.rep = n.rep, N = N, beta = beta, alpha = alpha,
psi.RE = psi.RE, p.RE = p.RE, sp = TRUE, sigma.sq = sigma.sq,
phi = phi, nu = nu, cov.model = 'matern')
pred.indx <- sample(1:J, round(J * .25), replace = FALSE)
y <- dat$y[, -pred.indx, , drop = FALSE]
# Occupancy covariates
X <- dat$X[-pred.indx, , drop = FALSE]
coords <- as.matrix(dat$coords[-pred.indx, , drop = FALSE])
# Prediction covariates
X.0 <- dat$X[pred.indx, , drop = FALSE]
coords.0 <- as.matrix(dat$coords[pred.indx, , drop = FALSE])
# Detection covariates
X.p <- dat$X.p[-pred.indx, , , drop = FALSE]
occ.covs <- X
colnames(occ.covs) <- c('int', 'occ.cov.1', 'occ.cov.2')
det.covs <- list(det.cov.1 = X.p[, , 2],
occ.cov.1 = X[, 2])
data.list <- list(y = y, coords = coords, occ.covs = occ.covs, det.covs = det.covs)
# Priors
prior.list <- list(beta.comm.normal = list(mean = 0, var = 2.72),
alpha.comm.normal = list(mean = 0, var = 2.72),
tau.sq.beta.ig = list(a = 0.1, b = 0.1),
tau.sq.alpha.ig = list(a = 0.1, b = 0.1),
nu.unif = list(0.5, 2.5))
# Starting values
inits.list <- list(alpha.comm = 0,
beta.comm = 0,
beta = 0,
alpha = 0,
tau.sq.beta = 1,
tau.sq.alpha = 1,
z = apply(y, c(1, 2), max, na.rm = TRUE))
# Tuning
tuning.list <- list(phi = 1, nu = 0.25)
batch.length <- 25
n.batch <- 10
n.report <- 100
occ.formula <- ~ 1
det.formula <- ~ occ.cov.1
out <- spMsPGOcc(occ.formula = occ.formula,
det.formula = det.formula,
data = data.list,
inits = inits.list,
n.batch = n.batch,
batch.length = batch.length,
accept.rate = 0.43,
priors = prior.list,
cov.model = "matern",
tuning = tuning.list,
n.omp.threads = 1,
verbose = FALSE,
NNGP = TRUE,
n.neighbors = 5,
search.type = 'cb',
n.report = 10,
n.burn = 100,
n.thin = 2,
n.chains = 2,
k.fold = 2,
k.fold.threads = 1)
# To make sure it worked --------------
test_that("out is of class spMsPGOcc", {
expect_s3_class(out, "spMsPGOcc")
})
# Check cross-validation --------------
test_that("cross-validation works", {
expect_equal(length(out$k.fold.deviance), N)
expect_type(out$k.fold.deviance, "double")
expect_equal(sum(out$k.fold.deviance < 0), 0)
})
# Check random effects ----------------
test_that("random effects are correct", {
expect_equal(out$pRE, FALSE)
expect_equal(out$psiRE, FALSE)
})
# Check output data output is correct -
test_that("out$y == y", {
expect_equal(out$y, y)
})
# Check missing values ----------------
test_that("missing value error handling works", {
# tmp.data <- data.list
# tmp.data$occ.covs[3, ] <- NA
# expect_error(spMsPGOcc(occ.formula = occ.formula,
# det.formula = det.formula,
# data = tmp.data,
# n.batch = 40,
# batch.length = batch.length,
# cov.model = "exponential",
# tuning = tuning.list,
# NNGP = TRUE,
# verbose = FALSE,
# n.neighbors = 5,
# search.type = 'cb',
# n.report = 10,
# n.burn = 100,
# n.chains = 1))
tmp.data <- data.list
tmp.data$y[1, 1, 1] <- 1
tmp.data$det.covs[[2]][1] <- NA
expect_error(spMsPGOcc(occ.formula = occ.formula,
det.formula = det.formula,
data = tmp.data,
n.batch = 40,
batch.length = batch.length,
cov.model = "exponential",
tuning = tuning.list,
NNGP = TRUE,
verbose = FALSE,
n.neighbors = 5,
search.type = 'cb',
n.report = 10,
n.burn = 100,
n.chains = 1))
tmp.data <- data.list
tmp.data$y[, 1, 1] <- NA
out <- spMsPGOcc(occ.formula = occ.formula,
det.formula = det.formula,
data = tmp.data,
n.batch = 40,
batch.length = batch.length,
cov.model = "exponential",
tuning = tuning.list,
NNGP = TRUE,
verbose = FALSE,
n.neighbors = 5,
search.type = 'cb',
n.report = 10,
n.burn = 100,
n.chains = 1)
expect_s3_class(out, "spMsPGOcc")
})
# Check default values ----------------
test_that("default priors, inits, burn, thin work", {
out <- spMsPGOcc(occ.formula = occ.formula,
det.formula = det.formula,
data = data.list,
n.batch = n.batch,
batch.length = batch.length,
accept.rate = 0.43,
cov.model = "exponential",
tuning = tuning.list,
NNGP = TRUE,
verbose = FALSE,
n.neighbors = 5,
n.chains = 1)
expect_s3_class(out, "spMsPGOcc")
})
test_that("all correlation functions work", {
out <- spMsPGOcc(occ.formula = occ.formula,
det.formula = det.formula,
data = data.list,
inits = inits.list,
n.batch = n.batch,
batch.length = batch.length,
accept.rate = 0.43,
priors = prior.list,
cov.model = "gaussian",
tuning = list(phi = 0.3),
n.omp.threads = 1,
verbose = FALSE,
NNGP = TRUE,
n.neighbors = 5,
search.type = 'cb',
n.report = 10,
n.burn = 100,
n.thin = 1,
n.chains = 1)
expect_s3_class(out, "spMsPGOcc")
out <- spMsPGOcc(occ.formula = occ.formula,
det.formula = det.formula,
data = data.list,
inits = inits.list,
n.batch = n.batch,
batch.length = batch.length,
accept.rate = 0.43,
priors = prior.list,
cov.model = "spherical",
tuning = list(phi = 0.3),
n.omp.threads = 1,
verbose = FALSE,
NNGP = TRUE,
n.neighbors = 5,
search.type = 'cb',
n.report = 10,
n.burn = 100,
n.thin = 1,
n.chains = 1)
expect_s3_class(out, "spMsPGOcc")
out <- spMsPGOcc(occ.formula = occ.formula,
det.formula = det.formula,
data = data.list,
inits = inits.list,
n.batch = n.batch,
batch.length = batch.length,
accept.rate = 0.43,
priors = list(nu.unif = list(0.4, 3)),
cov.model = "matern",
tuning = list(phi = 0.3, nu = 0.2),
n.omp.threads = 1,
verbose = FALSE,
NNGP = TRUE,
n.neighbors = 5,
search.type = 'cb',
n.report = 10,
n.burn = 100,
n.thin = 1,
n.chains = 1)
expect_s3_class(out, "spMsPGOcc")
})
test_that("verbose prints to the screen", {
expect_output(spMsPGOcc(occ.formula = occ.formula,
det.formula = det.formula,
data = data.list,
inits = inits.list,
n.batch = n.batch,
batch.length = batch.length,
accept.rate = 0.43,
priors = prior.list,
cov.model = "exponential",
tuning = tuning.list,
n.omp.threads = 1,
verbose = TRUE,
NNGP = TRUE,
n.neighbors = 5,
search.type = 'cb',
n.report = 10,
n.burn = 100,
n.thin = 1,
n.chains = 1))
})
# Check waicOcc -----------------------
test_that("waicOCC works for spMsPGOcc", {
# as.vector gets rid of names
waic.out <- as.vector(waicOcc(out))
expect_equal(length(waic.out), 3)
expect_equal(waic.out[3], -2 * (waic.out[1] - waic.out[2]))
})
test_that("fitted works for spMsPGOcc", {
fitted.out <- fitted(out)
expect_equal(length(fitted.out), 2)
expect_equal(class(fitted.out$y.rep.samples), "array")
expect_equal(class(fitted.out$p.samples), "array")
expect_equal(dim(fitted.out$y.rep.samples), dim(fitted.out$p.samples))
})
test_that("predict works for spMsPGOcc", {
n.post.samples <- out$n.post * out$n.chains
X.0 <- X.0[, 1, drop = FALSE]
colnames(X.0) <- c('int')
pred.out <- predict(out, X.0, coords.0, verbose = FALSE)
expect_type(pred.out, "list")
expect_equal(dim(pred.out$psi.0.samples), c(n.post.samples, N, nrow(X.0)))
expect_equal(dim(pred.out$z.0.samples), c(n.post.samples, N, nrow(X.0)))
})
test_that("detection prediction works", {
X.p.0 <- cbind(1, data.list$det.covs$occ.cov.1)
pred.out <- predict(out, X.p.0, type = 'detection')
expect_type(pred.out, 'list')
expect_equal(dim(pred.out$p.0.samples), c(out$n.post * out$n.chains, N, nrow(X.p.0)))
})
test_that("posterior predictive checks work for msPGOcc", {
n.post.samples <- out$n.post * out$n.chains
J.fit <- nrow(X)
ppc.out <- ppcOcc(out, 'chi-square', 2)
expect_type(ppc.out, "list")
expect_equal(dim(ppc.out$fit.y), c(n.post.samples, N))
expect_equal(dim(ppc.out$fit.y.rep), c(n.post.samples, N))
expect_equal(dim(ppc.out$fit.y.group.quants), c(5, N, max(n.rep)))
expect_equal(dim(ppc.out$fit.y.rep.group.quants), c(5, N, max(n.rep)))
ppc.out <- ppcOcc(out, 'chi-square', 1)
expect_type(ppc.out, "list")
expect_equal(dim(ppc.out$fit.y), c(n.post.samples, N))
expect_equal(dim(ppc.out$fit.y.rep), c(n.post.samples, N))
expect_equal(dim(ppc.out$fit.y.group.quants), c(5, N, J.fit))
expect_equal(dim(ppc.out$fit.y.rep.group.quants), c(5, N, J.fit))
ppc.out <- ppcOcc(out, 'freeman-tukey', 1)
expect_type(ppc.out, "list")
expect_equal(dim(ppc.out$fit.y), c(n.post.samples, N))
expect_equal(dim(ppc.out$fit.y.rep), c(n.post.samples, N))
expect_equal(dim(ppc.out$fit.y.group.quants), c(5, N, J.fit))
expect_equal(dim(ppc.out$fit.y.rep.group.quants), c(5, N, J.fit))
ppc.out <- ppcOcc(out, 'freeman-tukey', 2)
expect_type(ppc.out, "list")
expect_equal(dim(ppc.out$fit.y), c(n.post.samples, N))
expect_equal(dim(ppc.out$fit.y.rep), c(n.post.samples, N))
expect_equal(dim(ppc.out$fit.y.group.quants), c(5, N, max(n.rep)))
expect_equal(dim(ppc.out$fit.y.rep.group.quants), c(5, N, max(n.rep)))
})
# Random intercept on occurrence -------------------------------------------
J.x <- 8
J.y <- 8
J <- J.x * J.y
n.rep<- sample(2:4, size = J, replace = TRUE)
N <- 6
# Community-level covariate effects
# Occurrence
beta.mean <- c(0.2)
p.occ <- length(beta.mean)
tau.sq.beta <- c(0.6)
# Detection
alpha.mean <- c(0)
tau.sq.alpha <- c(1)
p.det <- length(alpha.mean)
# Random effects
psi.RE <- list(levels = c(45),
sigma.sq.psi = c(1.3))
p.RE <- list()
# Draw species-level effects from community means.
beta <- matrix(NA, nrow = N, ncol = p.occ)
alpha <- matrix(NA, nrow = N, ncol = p.det)
for (i in 1:p.occ) {
beta[, i] <- rnorm(N, beta.mean[i], sqrt(tau.sq.beta[i]))
}
for (i in 1:p.det) {
alpha[, i] <- rnorm(N, alpha.mean[i], sqrt(tau.sq.alpha[i]))
}
alpha.true <- alpha
phi <- rep(3 / .7, N)
sigma.sq <- rep(2, N)
nu <- rep(2, N)
dat <- simMsOcc(J.x = J.x, J.y = J.y, n.rep = n.rep, N = N, beta = beta, alpha = alpha,
psi.RE = psi.RE, p.RE = p.RE, sp = TRUE, sigma.sq = sigma.sq,
phi = phi, nu = nu, cov.model = 'matern')
pred.indx <- sample(1:J, round(J * .25), replace = FALSE)
y <- dat$y[, -pred.indx, , drop = FALSE]
# Occupancy covariates
X <- dat$X[-pred.indx, , drop = FALSE]
X.re <- dat$X.re[-pred.indx, , drop = FALSE]
coords <- as.matrix(dat$coords[-pred.indx, , drop = FALSE])
# Prediction covariates
X.0 <- dat$X[pred.indx, , drop = FALSE]
X.re.0 <- dat$X.re[pred.indx, , drop = FALSE]
coords.0 <- as.matrix(dat$coords[pred.indx, , drop = FALSE])
# Detection covariates
X.p <- dat$X.p[-pred.indx, , , drop = FALSE]
occ.covs <- cbind(X, X.re)
colnames(occ.covs) <- c('int', 'occ.factor.1')
data.list <- list(y = y, coords = coords, occ.covs = occ.covs)
# Priors
prior.list <- list(beta.comm.normal = list(mean = 0, var = 2.72),
alpha.comm.normal = list(mean = 0, var = 2.72),
tau.sq.beta.ig = list(a = 0.1, b = 0.1),
tau.sq.alpha.ig = list(a = 0.1, b = 0.1),
nu.unif = list(0.5, 2.5))
# Starting values
inits.list <- list(alpha.comm = 0,
beta.comm = 0,
beta = 0,
alpha = 0,
tau.sq.beta = 1,
tau.sq.alpha = 1,
z = apply(y, c(1, 2), max, na.rm = TRUE))
# Tuning
tuning.list <- list(phi = 1, nu = 0.25)
batch.length <- 25
n.batch <- 10
n.report <- 100
occ.formula <- ~ (1 | occ.factor.1)
det.formula <- ~ 1
out <- spMsPGOcc(occ.formula = occ.formula,
det.formula = det.formula,
data = data.list,
inits = inits.list,
n.batch = n.batch,
batch.length = batch.length,
accept.rate = 0.43,
priors = prior.list,
cov.model = "matern",
tuning = tuning.list,
n.omp.threads = 1,
verbose = FALSE,
NNGP = TRUE,
n.neighbors = 5,
search.type = 'cb',
n.report = 10,
n.burn = 100,
n.thin = 2,
n.chains = 2,
k.fold = 2,
k.fold.threads = 1)
# To make sure it worked --------------
test_that("out is of class spMsPGOcc", {
expect_s3_class(out, "spMsPGOcc")
})
# Check cross-validation --------------
test_that("cross-validation works", {
expect_equal(length(out$k.fold.deviance), N)
expect_type(out$k.fold.deviance, "double")
expect_equal(sum(out$k.fold.deviance < 0), 0)
})
# Check random effects ----------------
test_that("random effects are correct", {
expect_equal(out$pRE, FALSE)
expect_equal(out$psiRE, TRUE)
})
# Check RE error ----------------------
test_that("random effect gives error when non-numeric", {
data.list$occ.covs <- as.data.frame(data.list$occ.covs)
data.list$occ.covs$occ.factor.1 <- factor(data.list$occ.covs$occ.factor.1)
# data.list$det.covs$det.factor.1 <- factor(data.list$det.covs$det.factor.1)
expect_error(out <- spMsPGOcc(occ.formula = occ.formula,
det.formula = det.formula,
data = data.list,
inits = inits.list,
n.batch = n.batch,
batch.length = batch.length,
accept.rate = 0.43,
priors = prior.list,
cov.model = "matern",
tuning = tuning.list,
n.omp.threads = 1,
verbose = FALSE,
NNGP = TRUE,
n.neighbors = 5,
search.type = 'cb',
n.report = 10,
n.burn = 100,
n.thin = 2,
n.chains = 1))
data.list$occ.covs$occ.factor.1 <- as.character(factor(data.list$occ.covs$occ.factor.1))
# data.list$det.covs$det.factor.1 <- as.character(factor(data.list$det.covs$det.factor.1))
expect_error(out <- spMsPGOcc(occ.formula = occ.formula,
det.formula = det.formula,
data = data.list,
inits = inits.list,
n.batch = n.batch,
batch.length = batch.length,
accept.rate = 0.43,
priors = prior.list,
cov.model = "matern",
tuning = tuning.list,
n.omp.threads = 1,
verbose = FALSE,
NNGP = TRUE,
n.neighbors = 5,
search.type = 'cb',
n.report = 10,
n.burn = 100,
n.thin = 2,
n.chains = 1))
})
# Check RE levels ---------------------
test_that("random effect levels are correct", {
expect_equal(sort(unique(unlist(out$re.level.names))),
sort(unique(c(X.re))))
})
# Check output data output is correct -
test_that("out$y == y", {
expect_equal(out$y, y)
})
# Check missing values ----------------
test_that("missing value error handling works", {
tmp.data <- data.list
tmp.data$occ.covs[3, ] <- NA
expect_error(spMsPGOcc(occ.formula = occ.formula,
det.formula = det.formula,
data = tmp.data,
n.batch = 40,
batch.length = batch.length,
cov.model = "exponential",
tuning = tuning.list,
NNGP = TRUE,
verbose = FALSE,
n.neighbors = 5,
search.type = 'cb',
n.report = 10,
n.burn = 100,
n.chains = 1))
# tmp.data <- data.list
# tmp.data$det.covs[[1]][1] <- NA
# expect_error(spMsPGOcc(occ.formula = occ.formula,
# det.formula = det.formula,
# data = tmp.data,
# n.batch = 40,
# batch.length = batch.length,
# cov.model = "exponential",
# tuning = tuning.list,
# NNGP = TRUE,
# verbose = FALSE,
# n.neighbors = 5,
# search.type = 'cb',
# n.report = 10,
# n.burn = 100,
# n.chains = 1))
# tmp.data <- data.list
# tmp.data$y[, 1, 1] <- NA
# out <- spMsPGOcc(occ.formula = occ.formula,
# det.formula = det.formula,
# data = tmp.data,
# n.batch = 40,
# batch.length = batch.length,
# cov.model = "exponential",
# tuning = tuning.list,
# NNGP = TRUE,
# verbose = FALSE,
# n.neighbors = 5,
# search.type = 'cb',
# n.report = 10,
# n.burn = 100,
# n.chains = 1)
# expect_s3_class(out, "spMsPGOcc")
})
# Check default values ----------------
test_that("default priors, inits, burn, thin work", {
out <- spMsPGOcc(occ.formula = occ.formula,
det.formula = det.formula,
data = data.list,
n.batch = n.batch,
batch.length = batch.length,
accept.rate = 0.43,
cov.model = "exponential",
tuning = tuning.list,
NNGP = TRUE,
verbose = FALSE,
n.neighbors = 5,
n.chains = 1)
expect_s3_class(out, "spMsPGOcc")
})
test_that("all correlation functions work", {
out <- spMsPGOcc(occ.formula = occ.formula,
det.formula = det.formula,
data = data.list,
inits = inits.list,
n.batch = n.batch,
batch.length = batch.length,
accept.rate = 0.43,
priors = prior.list,
cov.model = "gaussian",
tuning = list(phi = 0.3),
n.omp.threads = 1,
verbose = FALSE,
NNGP = TRUE,
n.neighbors = 5,
search.type = 'cb',
n.report = 10,
n.burn = 100,
n.thin = 1,
n.chains = 1)
expect_s3_class(out, "spMsPGOcc")
out <- spMsPGOcc(occ.formula = occ.formula,
det.formula = det.formula,
data = data.list,
inits = inits.list,
n.batch = n.batch,
batch.length = batch.length,
accept.rate = 0.43,
priors = prior.list,
cov.model = "spherical",
tuning = list(phi = 0.3),
n.omp.threads = 1,
verbose = FALSE,
NNGP = TRUE,
n.neighbors = 5,
search.type = 'cb',
n.report = 10,
n.burn = 100,
n.thin = 1,
n.chains = 1)
expect_s3_class(out, "spMsPGOcc")
out <- spMsPGOcc(occ.formula = occ.formula,
det.formula = det.formula,
data = data.list,
inits = inits.list,
n.batch = n.batch,
batch.length = batch.length,
accept.rate = 0.43,
priors = list(nu.unif = list(0.4, 3)),
cov.model = "matern",
tuning = list(phi = 0.3, nu = 0.2),
n.omp.threads = 1,
verbose = FALSE,
NNGP = TRUE,
n.neighbors = 5,
search.type = 'cb',
n.report = 10,
n.burn = 100,
n.thin = 1,
n.chains = 1)
expect_s3_class(out, "spMsPGOcc")
})
test_that("verbose prints to the screen", {
expect_output(spMsPGOcc(occ.formula = occ.formula,
det.formula = det.formula,
data = data.list,
inits = inits.list,
n.batch = n.batch,
batch.length = batch.length,
accept.rate = 0.43,
priors = prior.list,
cov.model = "exponential",
tuning = tuning.list,
n.omp.threads = 1,
verbose = TRUE,
NNGP = TRUE,
n.neighbors = 5,
search.type = 'cb',
n.report = 10,
n.burn = 100,
n.thin = 1,
n.chains = 1))
})
# Check waicOcc -----------------------
test_that("waicOCC works for spMsPGOcc", {
# as.vector gets rid of names
waic.out <- as.vector(waicOcc(out))
expect_equal(length(waic.out), 3)
expect_equal(waic.out[3], -2 * (waic.out[1] - waic.out[2]))
})
test_that("fitted works for spMsPGOcc", {
fitted.out <- fitted(out)
expect_equal(length(fitted.out), 2)
expect_equal(class(fitted.out$y.rep.samples), "array")
expect_equal(class(fitted.out$p.samples), "array")
expect_equal(dim(fitted.out$y.rep.samples), dim(fitted.out$p.samples))
})
test_that("predict works for spMsPGOcc", {
n.post.samples <- out$n.post * out$n.chains
X.0 <- cbind(X.0, X.re.0)
colnames(X.0) <- c('int', 'occ.factor.1')
pred.out <- predict(out, X.0, coords.0, verbose = FALSE)
expect_type(pred.out, "list")
expect_equal(dim(pred.out$psi.0.samples), c(n.post.samples, N, nrow(X.0)))
expect_equal(dim(pred.out$z.0.samples), c(n.post.samples, N, nrow(X.0)))
})
test_that("detection prediction works", {
J.str <- 100
X.p.0 <- matrix(1, nrow = J.str, ncol = p.det)
pred.out <- predict(out, X.p.0, type = 'detection')
expect_type(pred.out, 'list')
expect_equal(dim(pred.out$p.0.samples), c(out$n.post * out$n.chains, N, J.str))
})
test_that("posterior predictive checks work for msPGOcc", {
n.post.samples <- out$n.post * out$n.chains
J.fit <- nrow(X)
ppc.out <- ppcOcc(out, 'chi-square', 2)
expect_type(ppc.out, "list")
expect_equal(dim(ppc.out$fit.y), c(n.post.samples, N))
expect_equal(dim(ppc.out$fit.y.rep), c(n.post.samples, N))
expect_equal(dim(ppc.out$fit.y.group.quants), c(5, N, max(n.rep)))
expect_equal(dim(ppc.out$fit.y.rep.group.quants), c(5, N, max(n.rep)))
ppc.out <- ppcOcc(out, 'chi-square', 1)
expect_type(ppc.out, "list")
expect_equal(dim(ppc.out$fit.y), c(n.post.samples, N))
expect_equal(dim(ppc.out$fit.y.rep), c(n.post.samples, N))
expect_equal(dim(ppc.out$fit.y.group.quants), c(5, N, J.fit))
expect_equal(dim(ppc.out$fit.y.rep.group.quants), c(5, N, J.fit))
ppc.out <- ppcOcc(out, 'freeman-tukey', 1)
expect_type(ppc.out, "list")
expect_equal(dim(ppc.out$fit.y), c(n.post.samples, N))
expect_equal(dim(ppc.out$fit.y.rep), c(n.post.samples, N))
expect_equal(dim(ppc.out$fit.y.group.quants), c(5, N, J.fit))
expect_equal(dim(ppc.out$fit.y.rep.group.quants), c(5, N, J.fit))
ppc.out <- ppcOcc(out, 'freeman-tukey', 2)
expect_type(ppc.out, "list")
expect_equal(dim(ppc.out$fit.y), c(n.post.samples, N))
expect_equal(dim(ppc.out$fit.y.rep), c(n.post.samples, N))
expect_equal(dim(ppc.out$fit.y.group.quants), c(5, N, max(n.rep)))
expect_equal(dim(ppc.out$fit.y.rep.group.quants), c(5, N, max(n.rep)))
})
# Multiple random intercepts on occurrence --------------------------------
J.x <- 8
J.y <- 8
J <- J.x * J.y
n.rep<- sample(2:4, size = J, replace = TRUE)
N <- 6
# Community-level covariate effects
# Occurrence
beta.mean <- c(0.2)
p.occ <- length(beta.mean)
tau.sq.beta <- c(0.6)
# Detection
alpha.mean <- c(0)
tau.sq.alpha <- c(1)
p.det <- length(alpha.mean)
# Random effects
psi.RE <- list(levels = c(45, 20),
sigma.sq.psi = c(1.3, 3.4))
p.RE <- list()
# Draw species-level effects from community means.
beta <- matrix(NA, nrow = N, ncol = p.occ)
alpha <- matrix(NA, nrow = N, ncol = p.det)
for (i in 1:p.occ) {
beta[, i] <- rnorm(N, beta.mean[i], sqrt(tau.sq.beta[i]))
}
for (i in 1:p.det) {
alpha[, i] <- rnorm(N, alpha.mean[i], sqrt(tau.sq.alpha[i]))
}
alpha.true <- alpha
phi <- rep(3 / .7, N)
sigma.sq <- rep(2, N)
nu <- rep(2, N)
dat <- simMsOcc(J.x = J.x, J.y = J.y, n.rep = n.rep, N = N, beta = beta, alpha = alpha,
psi.RE = psi.RE, p.RE = p.RE, sp = TRUE, sigma.sq = sigma.sq,
phi = phi, nu = nu, cov.model = 'matern')
pred.indx <- sample(1:J, round(J * .25), replace = FALSE)
y <- dat$y[, -pred.indx, , drop = FALSE]
# Occupancy covariates
X <- dat$X[-pred.indx, , drop = FALSE]
X.re <- dat$X.re[-pred.indx, , drop = FALSE]
coords <- as.matrix(dat$coords[-pred.indx, , drop = FALSE])
# Prediction covariates
X.0 <- dat$X[pred.indx, , drop = FALSE]
X.re.0 <- dat$X.re[pred.indx, , drop = FALSE]
coords.0 <- as.matrix(dat$coords[pred.indx, , drop = FALSE])
# Detection covariates
X.p <- dat$X.p[-pred.indx, , , drop = FALSE]
occ.covs <- cbind(X, X.re)
colnames(occ.covs) <- c('int', 'occ.factor.1', 'occ.factor.2')
data.list <- list(y = y, coords = coords, occ.covs = occ.covs)
# Priors
prior.list <- list(beta.comm.normal = list(mean = 0, var = 2.72),
alpha.comm.normal = list(mean = 0, var = 2.72),
tau.sq.beta.ig = list(a = 0.1, b = 0.1),
tau.sq.alpha.ig = list(a = 0.1, b = 0.1),
nu.unif = list(0.5, 2.5))
# Starting values
inits.list <- list(alpha.comm = 0,
beta.comm = 0,
beta = 0,
alpha = 0,
tau.sq.beta = 1,
tau.sq.alpha = 1,
z = apply(y, c(1, 2), max, na.rm = TRUE))
# Tuning
tuning.list <- list(phi = 1, nu = 0.25)
batch.length <- 25
n.batch <- 10
n.report <- 100
occ.formula <- ~ (1 | occ.factor.1) + (1 | occ.factor.2)
det.formula <- ~ 1
out <- spMsPGOcc(occ.formula = occ.formula,
det.formula = det.formula,
data = data.list,
inits = inits.list,
n.batch = n.batch,
batch.length = batch.length,
accept.rate = 0.43,
priors = prior.list,
cov.model = "matern",
tuning = tuning.list,
n.omp.threads = 1,
verbose = FALSE,
NNGP = TRUE,
n.neighbors = 5,
search.type = 'cb',
n.report = 10,
n.burn = 100,
n.thin = 2,
n.chains = 2,
k.fold = 2,
k.fold.threads = 1)
# To make sure it worked --------------
test_that("out is of class spMsPGOcc", {
expect_s3_class(out, "spMsPGOcc")
})
# Check cross-validation --------------
test_that("cross-validation works", {
expect_equal(length(out$k.fold.deviance), N)
expect_type(out$k.fold.deviance, "double")
expect_equal(sum(out$k.fold.deviance < 0), 0)
})
# Check random effects ----------------
test_that("random effects are correct", {
expect_equal(out$pRE, FALSE)
expect_equal(out$psiRE, TRUE)
})
# Check RE error ----------------------
test_that("random effect gives error when non-numeric", {
data.list$occ.covs <- as.data.frame(data.list$occ.covs)
data.list$occ.covs$occ.factor.1 <- factor(data.list$occ.covs$occ.factor.1)
# data.list$det.covs$det.factor.1 <- factor(data.list$det.covs$det.factor.1)
expect_error(out <- spMsPGOcc(occ.formula = occ.formula,
det.formula = det.formula,
data = data.list,
inits = inits.list,
n.batch = n.batch,
batch.length = batch.length,
accept.rate = 0.43,
priors = prior.list,
cov.model = "matern",
tuning = tuning.list,
n.omp.threads = 1,
verbose = FALSE,
NNGP = TRUE,
n.neighbors = 5,
search.type = 'cb',
n.report = 10,
n.burn = 100,
n.thin = 2,
n.chains = 1))
data.list$occ.covs$occ.factor.1 <- as.character(factor(data.list$occ.covs$occ.factor.1))
# data.list$det.covs$det.factor.1 <- as.character(factor(data.list$det.covs$det.factor.1))
expect_error(out <- spMsPGOcc(occ.formula = occ.formula,
det.formula = det.formula,
data = data.list,
inits = inits.list,
n.batch = n.batch,
batch.length = batch.length,
accept.rate = 0.43,
priors = prior.list,
cov.model = "matern",
tuning = tuning.list,
n.omp.threads = 1,
verbose = FALSE,
NNGP = TRUE,
n.neighbors = 5,
search.type = 'cb',
n.report = 10,
n.burn = 100,
n.thin = 2,
n.chains = 1))
})
# Check RE levels ---------------------
test_that("random effect levels are correct", {
expect_equal(sort(unique(unlist(out$re.level.names))),
sort(unique(c(X.re))))
})
# Check output data output is correct -
test_that("out$y == y", {
expect_equal(out$y, y)
})
# Check missing values ----------------
test_that("missing value error handling works", {
tmp.data <- data.list
tmp.data$occ.covs[3, ] <- NA
expect_error(spMsPGOcc(occ.formula = occ.formula,
det.formula = det.formula,
data = tmp.data,
n.batch = 40,
batch.length = batch.length,
cov.model = "exponential",
tuning = tuning.list,
NNGP = TRUE,
verbose = FALSE,
n.neighbors = 5,
search.type = 'cb',
n.report = 10,
n.burn = 100,
n.chains = 1))
# tmp.data <- data.list
# tmp.data$det.covs[[1]][1] <- NA
# expect_error(spMsPGOcc(occ.formula = occ.formula,
# det.formula = det.formula,
# data = tmp.data,
# n.batch = 40,
# batch.length = batch.length,
# cov.model = "exponential",
# tuning = tuning.list,
# NNGP = TRUE,
# verbose = FALSE,
# n.neighbors = 5,
# search.type = 'cb',
# n.report = 10,
# n.burn = 100,
# n.chains = 1))
# tmp.data <- data.list
# tmp.data$y[, 1, 1] <- NA
# out <- spMsPGOcc(occ.formula = occ.formula,
# det.formula = det.formula,
# data = tmp.data,
# n.batch = 40,
# batch.length = batch.length,
# cov.model = "exponential",
# tuning = tuning.list,
# NNGP = TRUE,
# verbose = FALSE,
# n.neighbors = 5,
# search.type = 'cb',
# n.report = 10,
# n.burn = 100,
# n.chains = 1)
# expect_s3_class(out, "spMsPGOcc")
})
# Check default values ----------------
test_that("default priors, inits, burn, thin work", {
out <- spMsPGOcc(occ.formula = occ.formula,
det.formula = det.formula,
data = data.list,
n.batch = n.batch,
batch.length = batch.length,
accept.rate = 0.43,
cov.model = "exponential",
tuning = tuning.list,
NNGP = TRUE,
verbose = FALSE,
n.neighbors = 5,
n.chains = 1)
expect_s3_class(out, "spMsPGOcc")
})
test_that("all correlation functions work", {
out <- spMsPGOcc(occ.formula = occ.formula,
det.formula = det.formula,
data = data.list,
inits = inits.list,
n.batch = n.batch,
batch.length = batch.length,
accept.rate = 0.43,
priors = prior.list,
cov.model = "gaussian",
tuning = list(phi = 0.3),
n.omp.threads = 1,
verbose = FALSE,
NNGP = TRUE,
n.neighbors = 5,
search.type = 'cb',
n.report = 10,
n.burn = 100,
n.thin = 1,
n.chains = 1)
expect_s3_class(out, "spMsPGOcc")
out <- spMsPGOcc(occ.formula = occ.formula,
det.formula = det.formula,
data = data.list,
inits = inits.list,
n.batch = n.batch,
batch.length = batch.length,
accept.rate = 0.43,
priors = prior.list,
cov.model = "spherical",
tuning = list(phi = 0.3),
n.omp.threads = 1,
verbose = FALSE,
NNGP = TRUE,
n.neighbors = 5,
search.type = 'cb',
n.report = 10,
n.burn = 100,
n.thin = 1,
n.chains = 1)
expect_s3_class(out, "spMsPGOcc")
out <- spMsPGOcc(occ.formula = occ.formula,
det.formula = det.formula,
data = data.list,
inits = inits.list,
n.batch = n.batch,
batch.length = batch.length,
accept.rate = 0.43,
priors = list(nu.unif = list(0.4, 3)),
cov.model = "matern",
tuning = list(phi = 0.3, nu = 0.2),
n.omp.threads = 1,
verbose = FALSE,
NNGP = TRUE,
n.neighbors = 5,
search.type = 'cb',
n.report = 10,
n.burn = 100,
n.thin = 1,
n.chains = 1)
expect_s3_class(out, "spMsPGOcc")
})
test_that("verbose prints to the screen", {
expect_output(spMsPGOcc(occ.formula = occ.formula,
det.formula = det.formula,
data = data.list,
inits = inits.list,
n.batch = n.batch,
batch.length = batch.length,
accept.rate = 0.43,
priors = prior.list,
cov.model = "exponential",
tuning = tuning.list,
n.omp.threads = 1,
verbose = TRUE,
NNGP = TRUE,
n.neighbors = 5,
search.type = 'cb',
n.report = 10,
n.burn = 100,
n.thin = 1,
n.chains = 1))
})
# Check waicOcc -----------------------
test_that("waicOCC works for spMsPGOcc", {
# as.vector gets rid of names
waic.out <- as.vector(waicOcc(out))
expect_equal(length(waic.out), 3)
expect_equal(waic.out[3], -2 * (waic.out[1] - waic.out[2]))
})
test_that("fitted works for spMsPGOcc", {
fitted.out <- fitted(out)
expect_equal(length(fitted.out), 2)
expect_equal(class(fitted.out$y.rep.samples), "array")
expect_equal(class(fitted.out$p.samples), "array")
expect_equal(dim(fitted.out$y.rep.samples), dim(fitted.out$p.samples))
})
test_that("predict works for spMsPGOcc", {
n.post.samples <- out$n.post * out$n.chains
X.0 <- cbind(X.0, X.re.0)
colnames(X.0) <- c('int', 'occ.factor.1', 'occ.factor.2')
pred.out <- predict(out, X.0, coords.0, verbose = FALSE)
expect_type(pred.out, "list")
expect_equal(dim(pred.out$psi.0.samples), c(n.post.samples, N, nrow(X.0)))
expect_equal(dim(pred.out$z.0.samples), c(n.post.samples, N, nrow(X.0)))
})
test_that("detection prediction works", {
J.str <- 100
X.p.0 <- matrix(1, nrow = J.str, ncol = p.det)
pred.out <- predict(out, X.p.0, type = 'detection')
expect_type(pred.out, 'list')
expect_equal(dim(pred.out$p.0.samples), c(out$n.post * out$n.chains, N, J.str))
})
test_that("posterior predictive checks work for msPGOcc", {
n.post.samples <- out$n.post * out$n.chains
J.fit <- nrow(X)
ppc.out <- ppcOcc(out, 'chi-square', 2)
expect_type(ppc.out, "list")
expect_equal(dim(ppc.out$fit.y), c(n.post.samples, N))
expect_equal(dim(ppc.out$fit.y.rep), c(n.post.samples, N))
expect_equal(dim(ppc.out$fit.y.group.quants), c(5, N, max(n.rep)))
expect_equal(dim(ppc.out$fit.y.rep.group.quants), c(5, N, max(n.rep)))
ppc.out <- ppcOcc(out, 'chi-square', 1)
expect_type(ppc.out, "list")
expect_equal(dim(ppc.out$fit.y), c(n.post.samples, N))
expect_equal(dim(ppc.out$fit.y.rep), c(n.post.samples, N))
expect_equal(dim(ppc.out$fit.y.group.quants), c(5, N, J.fit))
expect_equal(dim(ppc.out$fit.y.rep.group.quants), c(5, N, J.fit))
ppc.out <- ppcOcc(out, 'freeman-tukey', 1)
expect_type(ppc.out, "list")
expect_equal(dim(ppc.out$fit.y), c(n.post.samples, N))
expect_equal(dim(ppc.out$fit.y.rep), c(n.post.samples, N))
expect_equal(dim(ppc.out$fit.y.group.quants), c(5, N, J.fit))
expect_equal(dim(ppc.out$fit.y.rep.group.quants), c(5, N, J.fit))
ppc.out <- ppcOcc(out, 'freeman-tukey', 2)
expect_type(ppc.out, "list")
expect_equal(dim(ppc.out$fit.y), c(n.post.samples, N))
expect_equal(dim(ppc.out$fit.y.rep), c(n.post.samples, N))
expect_equal(dim(ppc.out$fit.y.group.quants), c(5, N, max(n.rep)))
expect_equal(dim(ppc.out$fit.y.rep.group.quants), c(5, N, max(n.rep)))
})
# Occurrence REs + covariates ---------------------------------------------
J.x <- 8
J.y <- 8
J <- J.x * J.y
n.rep<- sample(2:4, size = J, replace = TRUE)
N <- 6
# Community-level covariate effects
# Occurrence
beta.mean <- c(0.2, 1.2, -1.9)
p.occ <- length(beta.mean)
tau.sq.beta <- c(0.6, 1.5, 2.3)
# Detection
alpha.mean <- c(0)
tau.sq.alpha <- c(1)
p.det <- length(alpha.mean)
# Random effects
psi.RE <- list(levels = c(20, 10),
sigma.sq.psi = c(1.3, 3.4))
p.RE <- list()
# Draw species-level effects from community means.
beta <- matrix(NA, nrow = N, ncol = p.occ)
alpha <- matrix(NA, nrow = N, ncol = p.det)
for (i in 1:p.occ) {
beta[, i] <- rnorm(N, beta.mean[i], sqrt(tau.sq.beta[i]))
}
for (i in 1:p.det) {
alpha[, i] <- rnorm(N, alpha.mean[i], sqrt(tau.sq.alpha[i]))
}
alpha.true <- alpha
phi <- rep(3 / .7, N)
sigma.sq <- rep(2, N)
nu <- rep(2, N)
dat <- simMsOcc(J.x = J.x, J.y = J.y, n.rep = n.rep, N = N, beta = beta, alpha = alpha,
psi.RE = psi.RE, p.RE = p.RE, sp = TRUE, sigma.sq = sigma.sq,
phi = phi, nu = nu, cov.model = 'matern')
pred.indx <- sample(1:J, round(J * .25), replace = FALSE)
y <- dat$y[, -pred.indx, , drop = FALSE]
# Occupancy covariates
X <- dat$X[-pred.indx, , drop = FALSE]
X.re <- dat$X.re[-pred.indx, , drop = FALSE]
coords <- as.matrix(dat$coords[-pred.indx, , drop = FALSE])
# Prediction covariates
X.0 <- dat$X[pred.indx, , drop = FALSE]
X.re.0 <- dat$X.re[pred.indx, , drop = FALSE]
coords.0 <- as.matrix(dat$coords[pred.indx, , drop = FALSE])
# Detection covariates
X.p <- dat$X.p[-pred.indx, , , drop = FALSE]
occ.covs <- cbind(X, X.re)
colnames(occ.covs) <- c('int', 'occ.cov.1', 'occ.cov.2', 'occ.factor.1', 'occ.factor.2')
data.list <- list(y = y, coords = coords, occ.covs = occ.covs)
# Priors
prior.list <- list(beta.comm.normal = list(mean = 0, var = 2.72),
alpha.comm.normal = list(mean = 0, var = 2.72),
tau.sq.beta.ig = list(a = 0.1, b = 0.1),
tau.sq.alpha.ig = list(a = 0.1, b = 0.1),
nu.unif = list(0.5, 2.5))
# Starting values
inits.list <- list(alpha.comm = 0,
beta.comm = 0,
beta = 0,
alpha = 0,
tau.sq.beta = 1,
tau.sq.alpha = 1,
z = apply(y, c(1, 2), max, na.rm = TRUE))
# Tuning
tuning.list <- list(phi = 1, nu = 0.25)
batch.length <- 25
n.batch <- 10
n.report <- 100
occ.formula <- ~ occ.cov.1 + occ.cov.2 + (1 | occ.factor.1) + (1 | occ.factor.2)
det.formula <- ~ 1
out <- spMsPGOcc(occ.formula = occ.formula,
det.formula = det.formula,
data = data.list,
inits = inits.list,
n.batch = n.batch,
batch.length = batch.length,
accept.rate = 0.43,
priors = prior.list,
cov.model = "matern",
tuning = tuning.list,
n.omp.threads = 1,
verbose = FALSE,
NNGP = TRUE,
n.neighbors = 5,
search.type = 'cb',
n.report = 10,
n.burn = 100,
n.thin = 2,
n.chains = 2,
k.fold = 2,
k.fold.threads = 1)
# To make sure it worked --------------
test_that("out is of class spMsPGOcc", {
expect_s3_class(out, "spMsPGOcc")
})
# Check cross-validation --------------
test_that("cross-validation works", {
expect_equal(length(out$k.fold.deviance), N)
expect_type(out$k.fold.deviance, "double")
expect_equal(sum(out$k.fold.deviance < 0), 0)
})
# Check random effects ----------------
test_that("random effects are correct", {
expect_equal(out$pRE, FALSE)
expect_equal(out$psiRE, TRUE)
})
# Check RE error ----------------------
test_that("random effect gives error when non-numeric", {
data.list$occ.covs <- as.data.frame(data.list$occ.covs)
data.list$occ.covs$occ.factor.1 <- factor(data.list$occ.covs$occ.factor.1)
# data.list$det.covs$det.factor.1 <- factor(data.list$det.covs$det.factor.1)
expect_error(out <- spMsPGOcc(occ.formula = occ.formula,
det.formula = det.formula,
data = data.list,
inits = inits.list,
n.batch = n.batch,
batch.length = batch.length,
accept.rate = 0.43,
priors = prior.list,
cov.model = "matern",
tuning = tuning.list,
n.omp.threads = 1,
verbose = FALSE,
NNGP = TRUE,
n.neighbors = 5,
search.type = 'cb',
n.report = 10,
n.burn = 100,
n.thin = 2,
n.chains = 1))
data.list$occ.covs$occ.factor.1 <- as.character(factor(data.list$occ.covs$occ.factor.1))
# data.list$det.covs$det.factor.1 <- as.character(factor(data.list$det.covs$det.factor.1))
expect_error(out <- spMsPGOcc(occ.formula = occ.formula,
det.formula = det.formula,
data = data.list,
inits = inits.list,
n.batch = n.batch,
batch.length = batch.length,
accept.rate = 0.43,
priors = prior.list,
cov.model = "matern",
tuning = tuning.list,
n.omp.threads = 1,
verbose = FALSE,
NNGP = TRUE,
n.neighbors = 5,
search.type = 'cb',
n.report = 10,
n.burn = 100,
n.thin = 2,
n.chains = 1))
})
# Check RE levels ---------------------
test_that("random effect levels are correct", {
expect_equal(sort(unique(unlist(out$re.level.names))),
sort(unique(c(X.re))))
})
# Check output data output is correct -
test_that("out$y == y", {
expect_equal(out$y, y)
})
# Check missing values ----------------
test_that("missing value error handling works", {
tmp.data <- data.list
tmp.data$occ.covs[3, ] <- NA
expect_error(spMsPGOcc(occ.formula = occ.formula,
det.formula = det.formula,
data = tmp.data,
n.batch = 40,
batch.length = batch.length,
cov.model = "exponential",
tuning = tuning.list,
NNGP = TRUE,
verbose = FALSE,
n.neighbors = 5,
search.type = 'cb',
n.report = 10,
n.burn = 100,
n.chains = 1))
# tmp.data <- data.list
# tmp.data$det.covs[[1]][1] <- NA
# expect_error(spMsPGOcc(occ.formula = occ.formula,
# det.formula = det.formula,
# data = tmp.data,
# n.batch = 40,
# batch.length = batch.length,
# cov.model = "exponential",
# tuning = tuning.list,
# NNGP = TRUE,
# verbose = FALSE,
# n.neighbors = 5,
# search.type = 'cb',
# n.report = 10,
# n.burn = 100,
# n.chains = 1))
# tmp.data <- data.list
# tmp.data$y[, 1, 1] <- NA
# out <- spMsPGOcc(occ.formula = occ.formula,
# det.formula = det.formula,
# data = tmp.data,
# n.batch = 40,
# batch.length = batch.length,
# cov.model = "exponential",
# tuning = tuning.list,
# NNGP = TRUE,
# verbose = FALSE,
# n.neighbors = 5,
# search.type = 'cb',
# n.report = 10,
# n.burn = 100,
# n.chains = 1)
# expect_s3_class(out, "spMsPGOcc")
})
# Check default values ----------------
test_that("default priors, inits, burn, thin work", {
out <- spMsPGOcc(occ.formula = occ.formula,
det.formula = det.formula,
data = data.list,
n.batch = n.batch,
batch.length = batch.length,
accept.rate = 0.43,
cov.model = "exponential",
tuning = tuning.list,
NNGP = TRUE,
verbose = FALSE,
n.neighbors = 5,
n.chains = 1)
expect_s3_class(out, "spMsPGOcc")
})
test_that("all correlation functions work", {
out <- spMsPGOcc(occ.formula = occ.formula,
det.formula = det.formula,
data = data.list,
inits = inits.list,
n.batch = n.batch,
batch.length = batch.length,
accept.rate = 0.43,
priors = prior.list,
cov.model = "gaussian",
tuning = list(phi = 0.3),
n.omp.threads = 1,
verbose = FALSE,
NNGP = TRUE,
n.neighbors = 5,
search.type = 'cb',
n.report = 10,
n.burn = 100,
n.thin = 1,
n.chains = 1)
expect_s3_class(out, "spMsPGOcc")
out <- spMsPGOcc(occ.formula = occ.formula,
det.formula = det.formula,
data = data.list,
inits = inits.list,
n.batch = n.batch,
batch.length = batch.length,
accept.rate = 0.43,
priors = prior.list,
cov.model = "spherical",
tuning = list(phi = 0.3),
n.omp.threads = 1,
verbose = FALSE,
NNGP = TRUE,
n.neighbors = 5,
search.type = 'cb',
n.report = 10,
n.burn = 100,
n.thin = 1,
n.chains = 1)
expect_s3_class(out, "spMsPGOcc")
out <- spMsPGOcc(occ.formula = occ.formula,
det.formula = det.formula,
data = data.list,
inits = inits.list,
n.batch = n.batch,
batch.length = batch.length,
accept.rate = 0.43,
priors = list(nu.unif = list(0.4, 3)),
cov.model = "matern",
tuning = list(phi = 0.3, nu = 0.2),
n.omp.threads = 1,
verbose = FALSE,
NNGP = TRUE,
n.neighbors = 5,
search.type = 'cb',
n.report = 10,
n.burn = 100,
n.thin = 1,
n.chains = 1)
expect_s3_class(out, "spMsPGOcc")
})
test_that("verbose prints to the screen", {
expect_output(spMsPGOcc(occ.formula = occ.formula,
det.formula = det.formula,
data = data.list,
inits = inits.list,
n.batch = n.batch,
batch.length = batch.length,
accept.rate = 0.43,
priors = prior.list,
cov.model = "exponential",
tuning = tuning.list,
n.omp.threads = 1,
verbose = TRUE,
NNGP = TRUE,
n.neighbors = 5,
search.type = 'cb',
n.report = 10,
n.burn = 100,
n.thin = 1,
n.chains = 1))
})
# Check waicOcc -----------------------
test_that("waicOCC works for spMsPGOcc", {
# as.vector gets rid of names
waic.out <- as.vector(waicOcc(out))
expect_equal(length(waic.out), 3)
expect_equal(waic.out[3], -2 * (waic.out[1] - waic.out[2]))
})
test_that("fitted works for spMsPGOcc", {
fitted.out <- fitted(out)
expect_equal(length(fitted.out), 2)
expect_equal(class(fitted.out$y.rep.samples), "array")
expect_equal(class(fitted.out$p.samples), "array")
expect_equal(dim(fitted.out$y.rep.samples), dim(fitted.out$p.samples))
})
test_that("predict works for spMsPGOcc", {
n.post.samples <- out$n.post * out$n.chains
X.0 <- cbind(X.0, X.re.0)
colnames(X.0) <- c('int', 'occ.cov.1', 'occ.cov.2', 'occ.factor.1', 'occ.factor.2')
pred.out <- predict(out, X.0, coords.0, verbose = FALSE)
expect_type(pred.out, "list")
expect_equal(dim(pred.out$psi.0.samples), c(n.post.samples, N, nrow(X.0)))
expect_equal(dim(pred.out$z.0.samples), c(n.post.samples, N, nrow(X.0)))
})
test_that("detection prediction works", {
J.str <- 100
X.p.0 <- matrix(1, nrow = J.str, ncol = p.det)
pred.out <- predict(out, X.p.0, type = 'detection')
expect_type(pred.out, 'list')
expect_equal(dim(pred.out$p.0.samples), c(out$n.post * out$n.chains, N, J.str))
})
test_that("posterior predictive checks work for msPGOcc", {
n.post.samples <- out$n.post * out$n.chains
J.fit <- nrow(X)
ppc.out <- ppcOcc(out, 'chi-square', 2)
expect_type(ppc.out, "list")
expect_equal(dim(ppc.out$fit.y), c(n.post.samples, N))
expect_equal(dim(ppc.out$fit.y.rep), c(n.post.samples, N))
expect_equal(dim(ppc.out$fit.y.group.quants), c(5, N, max(n.rep)))
expect_equal(dim(ppc.out$fit.y.rep.group.quants), c(5, N, max(n.rep)))
ppc.out <- ppcOcc(out, 'chi-square', 1)
expect_type(ppc.out, "list")
expect_equal(dim(ppc.out$fit.y), c(n.post.samples, N))
expect_equal(dim(ppc.out$fit.y.rep), c(n.post.samples, N))
expect_equal(dim(ppc.out$fit.y.group.quants), c(5, N, J.fit))
expect_equal(dim(ppc.out$fit.y.rep.group.quants), c(5, N, J.fit))
ppc.out <- ppcOcc(out, 'freeman-tukey', 1)
expect_type(ppc.out, "list")
expect_equal(dim(ppc.out$fit.y), c(n.post.samples, N))
expect_equal(dim(ppc.out$fit.y.rep), c(n.post.samples, N))
expect_equal(dim(ppc.out$fit.y.group.quants), c(5, N, J.fit))
expect_equal(dim(ppc.out$fit.y.rep.group.quants), c(5, N, J.fit))
ppc.out <- ppcOcc(out, 'freeman-tukey', 2)
expect_type(ppc.out, "list")
expect_equal(dim(ppc.out$fit.y), c(n.post.samples, N))
expect_equal(dim(ppc.out$fit.y.rep), c(n.post.samples, N))
expect_equal(dim(ppc.out$fit.y.group.quants), c(5, N, max(n.rep)))
expect_equal(dim(ppc.out$fit.y.rep.group.quants), c(5, N, max(n.rep)))
})
# Occurrence REs + covariates in everything -------------------------------
J.x <- 8
J.y <- 8
J <- J.x * J.y
n.rep<- sample(2:4, size = J, replace = TRUE)
N <- 6
# Community-level covariate effects
# Occurrence
beta.mean <- c(0.2, 1.2, -1.9)
p.occ <- length(beta.mean)
tau.sq.beta <- c(0.6, 1.5, 2.3)
# Detection
alpha.mean <- c(0, 1.2)
tau.sq.alpha <- c(1, 2.3)
p.det <- length(alpha.mean)
# Random effects
psi.RE <- list(levels = c(20, 10),
sigma.sq.psi = c(1.3, 3.4))
p.RE <- list()
# Draw species-level effects from community means.
beta <- matrix(NA, nrow = N, ncol = p.occ)
alpha <- matrix(NA, nrow = N, ncol = p.det)
for (i in 1:p.occ) {
beta[, i] <- rnorm(N, beta.mean[i], sqrt(tau.sq.beta[i]))
}
for (i in 1:p.det) {
alpha[, i] <- rnorm(N, alpha.mean[i], sqrt(tau.sq.alpha[i]))
}
alpha.true <- alpha
phi <- rep(3 / .7, N)
sigma.sq <- rep(2, N)
nu <- rep(2, N)
dat <- simMsOcc(J.x = J.x, J.y = J.y, n.rep = n.rep, N = N, beta = beta, alpha = alpha,
psi.RE = psi.RE, p.RE = p.RE, sp = TRUE, sigma.sq = sigma.sq,
phi = phi, nu = nu, cov.model = 'matern')
pred.indx <- sample(1:J, round(J * .25), replace = FALSE)
y <- dat$y[, -pred.indx, , drop = FALSE]
# Occupancy covariates
X <- dat$X[-pred.indx, , drop = FALSE]
X.re <- dat$X.re[-pred.indx, , drop = FALSE]
coords <- as.matrix(dat$coords[-pred.indx, , drop = FALSE])
# Prediction covariates
X.0 <- dat$X[pred.indx, , drop = FALSE]
X.re.0 <- dat$X.re[pred.indx, , drop = FALSE]
coords.0 <- as.matrix(dat$coords[pred.indx, , drop = FALSE])
# Detection covariates
X.p <- dat$X.p[-pred.indx, , , drop = FALSE]
occ.covs <- cbind(X, X.re)
colnames(occ.covs) <- c('int', 'occ.cov.1', 'occ.cov.2', 'occ.factor.1', 'occ.factor.2')
det.covs <- list(det.cov.1 = X.p[, , 2])
data.list <- list(y = y, coords = coords, occ.covs = occ.covs, det.covs = det.covs)
# Priors
prior.list <- list(beta.comm.normal = list(mean = 0, var = 2.72),
alpha.comm.normal = list(mean = 0, var = 2.72),
tau.sq.beta.ig = list(a = 0.1, b = 0.1),
tau.sq.alpha.ig = list(a = 0.1, b = 0.1),
nu.unif = list(0.5, 2.5))
# Starting values
inits.list <- list(alpha.comm = 0,
beta.comm = 0,
beta = 0,
alpha = 0,
tau.sq.beta = 1,
tau.sq.alpha = 1,
z = apply(y, c(1, 2), max, na.rm = TRUE))
# Tuning
tuning.list <- list(phi = 1, nu = 0.25)
batch.length <- 25
n.batch <- 10
n.report <- 100
occ.formula <- ~ occ.cov.1 + occ.cov.2 + (1 | occ.factor.1) + (1 | occ.factor.2)
det.formula <- ~ det.cov.1
out <- spMsPGOcc(occ.formula = occ.formula,
det.formula = det.formula,
data = data.list,
inits = inits.list,
n.batch = n.batch,
batch.length = batch.length,
accept.rate = 0.43,
priors = prior.list,
cov.model = "matern",
tuning = tuning.list,
n.omp.threads = 1,
verbose = FALSE,
NNGP = TRUE,
n.neighbors = 5,
search.type = 'cb',
n.report = 10,
n.burn = 100,
n.thin = 2,
n.chains = 2,
k.fold = 2,
k.fold.threads = 1)
# To make sure it worked --------------
test_that("out is of class spMsPGOcc", {
expect_s3_class(out, "spMsPGOcc")
})
# Check cross-validation --------------
test_that("cross-validation works", {
expect_equal(length(out$k.fold.deviance), N)
expect_type(out$k.fold.deviance, "double")
expect_equal(sum(out$k.fold.deviance < 0), 0)
})
# Check random effects ----------------
test_that("random effects are correct", {
expect_equal(out$pRE, FALSE)
expect_equal(out$psiRE, TRUE)
})
# Check RE error ----------------------
test_that("random effect gives error when non-numeric", {
data.list$occ.covs <- as.data.frame(data.list$occ.covs)
data.list$occ.covs$occ.factor.1 <- factor(data.list$occ.covs$occ.factor.1)
# data.list$det.covs$det.factor.1 <- factor(data.list$det.covs$det.factor.1)
expect_error(out <- spMsPGOcc(occ.formula = occ.formula,
det.formula = det.formula,
data = data.list,
inits = inits.list,
n.batch = n.batch,
batch.length = batch.length,
accept.rate = 0.43,
priors = prior.list,
cov.model = "matern",
tuning = tuning.list,
n.omp.threads = 1,
verbose = FALSE,
NNGP = TRUE,
n.neighbors = 5,
search.type = 'cb',
n.report = 10,
n.burn = 100,
n.thin = 2,
n.chains = 1))
data.list$occ.covs$occ.factor.1 <- as.character(factor(data.list$occ.covs$occ.factor.1))
# data.list$det.covs$det.factor.1 <- as.character(factor(data.list$det.covs$det.factor.1))
expect_error(out <- spMsPGOcc(occ.formula = occ.formula,
det.formula = det.formula,
data = data.list,
inits = inits.list,
n.batch = n.batch,
batch.length = batch.length,
accept.rate = 0.43,
priors = prior.list,
cov.model = "matern",
tuning = tuning.list,
n.omp.threads = 1,
verbose = FALSE,
NNGP = TRUE,
n.neighbors = 5,
search.type = 'cb',
n.report = 10,
n.burn = 100,
n.thin = 2,
n.chains = 1))
})
# Check RE levels ---------------------
test_that("random effect levels are correct", {
expect_equal(sort(unique(unlist(out$re.level.names))),
sort(unique(c(X.re))))
})
# Check output data output is correct -
test_that("out$y == y", {
expect_equal(out$y, y)
})
# Check missing values ----------------
test_that("missing value error handling works", {
tmp.data <- data.list
tmp.data$occ.covs[3, ] <- NA
expect_error(spMsPGOcc(occ.formula = occ.formula,
det.formula = det.formula,
data = tmp.data,
n.batch = 40,
batch.length = batch.length,
cov.model = "exponential",
tuning = tuning.list,
NNGP = TRUE,
verbose = FALSE,
n.neighbors = 5,
search.type = 'cb',
n.report = 10,
n.burn = 100,
n.chains = 1))
tmp.data <- data.list
tmp.data$y[1, 1, 1] <- 1
tmp.data$det.covs[[1]][1] <- NA
expect_error(spMsPGOcc(occ.formula = occ.formula,
det.formula = det.formula,
data = tmp.data,
n.batch = 40,
batch.length = batch.length,
cov.model = "exponential",
tuning = tuning.list,
NNGP = TRUE,
verbose = FALSE,
n.neighbors = 5,
search.type = 'cb',
n.report = 10,
n.burn = 100,
n.chains = 1))
tmp.data <- data.list
tmp.data$y[, 1, 1] <- NA
out <- spMsPGOcc(occ.formula = occ.formula,
det.formula = det.formula,
data = tmp.data,
n.batch = 40,
batch.length = batch.length,
cov.model = "exponential",
tuning = tuning.list,
NNGP = TRUE,
verbose = FALSE,
n.neighbors = 5,
search.type = 'cb',
n.report = 10,
n.burn = 100,
n.chains = 1)
expect_s3_class(out, "spMsPGOcc")
})
# Check default values ----------------
test_that("default priors, inits, burn, thin work", {
out <- spMsPGOcc(occ.formula = occ.formula,
det.formula = det.formula,
data = data.list,
n.batch = n.batch,
batch.length = batch.length,
accept.rate = 0.43,
cov.model = "exponential",
tuning = tuning.list,
NNGP = TRUE,
verbose = FALSE,
n.neighbors = 5,
n.chains = 1)
expect_s3_class(out, "spMsPGOcc")
})
test_that("all correlation functions work", {
out <- spMsPGOcc(occ.formula = occ.formula,
det.formula = det.formula,
data = data.list,
inits = inits.list,
n.batch = n.batch,
batch.length = batch.length,
accept.rate = 0.43,
priors = prior.list,
cov.model = "gaussian",
tuning = list(phi = 0.3),
n.omp.threads = 1,
verbose = FALSE,
NNGP = TRUE,
n.neighbors = 5,
search.type = 'cb',
n.report = 10,
n.burn = 100,
n.thin = 1,
n.chains = 1)
expect_s3_class(out, "spMsPGOcc")
out <- spMsPGOcc(occ.formula = occ.formula,
det.formula = det.formula,
data = data.list,
inits = inits.list,
n.batch = n.batch,
batch.length = batch.length,
accept.rate = 0.43,
priors = prior.list,
cov.model = "spherical",
tuning = list(phi = 0.3),
n.omp.threads = 1,
verbose = FALSE,
NNGP = TRUE,
n.neighbors = 5,
search.type = 'cb',
n.report = 10,
n.burn = 100,
n.thin = 1,
n.chains = 1)
expect_s3_class(out, "spMsPGOcc")
out <- spMsPGOcc(occ.formula = occ.formula,
det.formula = det.formula,
data = data.list,
inits = inits.list,
n.batch = n.batch,
batch.length = batch.length,
accept.rate = 0.43,
priors = list(nu.unif = list(0.4, 3)),
cov.model = "matern",
tuning = list(phi = 0.3, nu = 0.2),
n.omp.threads = 1,
verbose = FALSE,
NNGP = TRUE,
n.neighbors = 5,
search.type = 'cb',
n.report = 10,
n.burn = 100,
n.thin = 1,
n.chains = 1)
expect_s3_class(out, "spMsPGOcc")
})
test_that("verbose prints to the screen", {
expect_output(spMsPGOcc(occ.formula = occ.formula,
det.formula = det.formula,
data = data.list,
inits = inits.list,
n.batch = n.batch,
batch.length = batch.length,
accept.rate = 0.43,
priors = prior.list,
cov.model = "exponential",
tuning = tuning.list,
n.omp.threads = 1,
verbose = TRUE,
NNGP = TRUE,
n.neighbors = 5,
search.type = 'cb',
n.report = 10,
n.burn = 100,
n.thin = 1,
n.chains = 1))
})
# Check waicOcc -----------------------
test_that("waicOCC works for spMsPGOcc", {
# as.vector gets rid of names
waic.out <- as.vector(waicOcc(out))
expect_equal(length(waic.out), 3)
expect_equal(waic.out[3], -2 * (waic.out[1] - waic.out[2]))
})
test_that("fitted works for spMsPGOcc", {
fitted.out <- fitted(out)
expect_equal(length(fitted.out), 2)
expect_equal(class(fitted.out$y.rep.samples), "array")
expect_equal(class(fitted.out$p.samples), "array")
expect_equal(dim(fitted.out$y.rep.samples), dim(fitted.out$p.samples))
})
test_that("predict works for spMsPGOcc", {
n.post.samples <- out$n.post * out$n.chains
X.0 <- cbind(X.0, X.re.0)
colnames(X.0) <- c('int', 'occ.cov.1', 'occ.cov.2', 'occ.factor.1', 'occ.factor.2')
pred.out <- predict(out, X.0, coords.0, verbose = FALSE)
expect_type(pred.out, "list")
expect_equal(dim(pred.out$psi.0.samples), c(n.post.samples, N, nrow(X.0)))
expect_equal(dim(pred.out$z.0.samples), c(n.post.samples, N, nrow(X.0)))
})
test_that("detection prediction works", {
X.p.0 <- dat$X.p[, 1, ]
pred.out <- predict(out, X.p.0, type = 'detection')
expect_type(pred.out, 'list')
expect_equal(dim(pred.out$p.0.samples), c(out$n.post * out$n.chains, N, J))
})
test_that("posterior predictive checks work for msPGOcc", {
n.post.samples <- out$n.post * out$n.chains
J.fit <- nrow(X)
ppc.out <- ppcOcc(out, 'chi-square', 2)
expect_type(ppc.out, "list")
expect_equal(dim(ppc.out$fit.y), c(n.post.samples, N))
expect_equal(dim(ppc.out$fit.y.rep), c(n.post.samples, N))
expect_equal(dim(ppc.out$fit.y.group.quants), c(5, N, max(n.rep)))
expect_equal(dim(ppc.out$fit.y.rep.group.quants), c(5, N, max(n.rep)))
ppc.out <- ppcOcc(out, 'chi-square', 1)
expect_type(ppc.out, "list")
expect_equal(dim(ppc.out$fit.y), c(n.post.samples, N))
expect_equal(dim(ppc.out$fit.y.rep), c(n.post.samples, N))
expect_equal(dim(ppc.out$fit.y.group.quants), c(5, N, J.fit))
expect_equal(dim(ppc.out$fit.y.rep.group.quants), c(5, N, J.fit))
ppc.out <- ppcOcc(out, 'freeman-tukey', 1)
expect_type(ppc.out, "list")
expect_equal(dim(ppc.out$fit.y), c(n.post.samples, N))
expect_equal(dim(ppc.out$fit.y.rep), c(n.post.samples, N))
expect_equal(dim(ppc.out$fit.y.group.quants), c(5, N, J.fit))
expect_equal(dim(ppc.out$fit.y.rep.group.quants), c(5, N, J.fit))
ppc.out <- ppcOcc(out, 'freeman-tukey', 2)
expect_type(ppc.out, "list")
expect_equal(dim(ppc.out$fit.y), c(n.post.samples, N))
expect_equal(dim(ppc.out$fit.y.rep), c(n.post.samples, N))
expect_equal(dim(ppc.out$fit.y.group.quants), c(5, N, max(n.rep)))
expect_equal(dim(ppc.out$fit.y.rep.group.quants), c(5, N, max(n.rep)))
})
# Random intercept on detection -------------------------------------------
J.x <- 8
J.y <- 8
J <- J.x * J.y
n.rep<- sample(2:4, size = J, replace = TRUE)
N <- 6
# Community-level covariate effects
# Occurrence
beta.mean <- c(0.2)
p.occ <- length(beta.mean)
tau.sq.beta <- c(0.6)
# Detection
alpha.mean <- c(0)
tau.sq.alpha <- c(1)
p.det <- length(alpha.mean)
# Random effects
psi.RE <- list()
p.RE <- list(levels = c(50),
sigma.sq.p = c(2.50))
# Draw species-level effects from community means.
beta <- matrix(NA, nrow = N, ncol = p.occ)
alpha <- matrix(NA, nrow = N, ncol = p.det)
for (i in 1:p.occ) {
beta[, i] <- rnorm(N, beta.mean[i], sqrt(tau.sq.beta[i]))
}
for (i in 1:p.det) {
alpha[, i] <- rnorm(N, alpha.mean[i], sqrt(tau.sq.alpha[i]))
}
alpha.true <- alpha
phi <- rep(3 / .7, N)
sigma.sq <- rep(2, N)
nu <- rep(2, N)
dat <- simMsOcc(J.x = J.x, J.y = J.y, n.rep = n.rep, N = N, beta = beta, alpha = alpha,
psi.RE = psi.RE, p.RE = p.RE, sp = TRUE, sigma.sq = sigma.sq,
phi = phi, nu = nu, cov.model = 'matern')
pred.indx <- sample(1:J, round(J * .25), replace = FALSE)
y <- dat$y[, -pred.indx, , drop = FALSE]
# Occupancy covariates
X <- dat$X[-pred.indx, , drop = FALSE]
# X.re <- dat$X.re[-pred.indx, , drop = FALSE]
coords <- as.matrix(dat$coords[-pred.indx, , drop = FALSE])
# Prediction covariates
X.0 <- dat$X[pred.indx, , drop = FALSE]
# X.re.0 <- dat$X.re[pred.indx, , drop = FALSE]
coords.0 <- as.matrix(dat$coords[pred.indx, , drop = FALSE])
# Detection covariates
X.p <- dat$X.p[-pred.indx, , , drop = FALSE]
X.p.re <- dat$X.p.re[-pred.indx, , , drop = FALSE]
occ.covs <- cbind(X)
colnames(occ.covs) <- c('int')
det.covs <- list(det.factor.1 = X.p.re[, , 1])
data.list <- list(y = y, coords = coords, occ.covs = occ.covs, det.covs = det.covs)
# Priors
prior.list <- list(beta.comm.normal = list(mean = 0, var = 2.72),
alpha.comm.normal = list(mean = 0, var = 2.72),
tau.sq.beta.ig = list(a = 0.1, b = 0.1),
tau.sq.alpha.ig = list(a = 0.1, b = 0.1),
nu.unif = list(0.5, 2.5))
# Starting values
inits.list <- list(alpha.comm = 0,
beta.comm = 0,
beta = 0,
alpha = 0,
tau.sq.beta = 1,
tau.sq.alpha = 1,
z = apply(y, c(1, 2), max, na.rm = TRUE))
# Tuning
tuning.list <- list(phi = 1, nu = 0.25)
batch.length <- 25
n.batch <- 10
n.report <- 100
occ.formula <- ~ 1
det.formula <- ~ (1 | det.factor.1)
out <- spMsPGOcc(occ.formula = occ.formula,
det.formula = det.formula,
data = data.list,
inits = inits.list,
n.batch = n.batch,
batch.length = batch.length,
accept.rate = 0.43,
priors = prior.list,
cov.model = "matern",
tuning = tuning.list,
n.omp.threads = 1,
verbose = FALSE,
NNGP = TRUE,
n.neighbors = 5,
search.type = 'cb',
n.report = 10,
n.burn = 100,
n.thin = 2,
n.chains = 2,
k.fold = 2,
k.fold.threads = 1)
# To make sure it worked --------------
test_that("out is of class spMsPGOcc", {
expect_s3_class(out, "spMsPGOcc")
})
# Check cross-validation --------------
test_that("cross-validation works", {
expect_equal(length(out$k.fold.deviance), N)
expect_type(out$k.fold.deviance, "double")
expect_equal(sum(out$k.fold.deviance < 0), 0)
})
# Check random effects ----------------
test_that("random effects are correct", {
expect_equal(out$pRE, TRUE)
expect_equal(out$psiRE, FALSE)
})
# Check RE error ----------------------
test_that("random effect gives error when non-numeric", {
# data.list$occ.covs <- as.data.frame(data.list$occ.covs)
# data.list$occ.covs$occ.factor.1 <- factor(data.list$occ.covs$occ.factor.1)
data.list$det.covs$det.factor.1 <- factor(data.list$det.covs$det.factor.1)
expect_error(out <- spMsPGOcc(occ.formula = occ.formula,
det.formula = det.formula,
data = data.list,
inits = inits.list,
n.batch = n.batch,
batch.length = batch.length,
accept.rate = 0.43,
priors = prior.list,
cov.model = "matern",
tuning = tuning.list,
n.omp.threads = 1,
verbose = FALSE,
NNGP = TRUE,
n.neighbors = 5,
search.type = 'cb',
n.report = 10,
n.burn = 100,
n.thin = 2,
n.chains = 1))
# data.list$occ.covs$occ.factor.1 <- as.character(factor(data.list$occ.covs$occ.factor.1))
data.list$det.covs$det.factor.1 <- as.character(factor(data.list$det.covs$det.factor.1))
expect_error(out <- spMsPGOcc(occ.formula = occ.formula,
det.formula = det.formula,
data = data.list,
inits = inits.list,
n.batch = n.batch,
batch.length = batch.length,
accept.rate = 0.43,
priors = prior.list,
cov.model = "matern",
tuning = tuning.list,
n.omp.threads = 1,
verbose = FALSE,
NNGP = TRUE,
n.neighbors = 5,
search.type = 'cb',
n.report = 10,
n.burn = 100,
n.thin = 2,
n.chains = 1))
})
# Check RE levels ---------------------
test_that("random effect levels are correct", {
expect_equal(sort(unique(unlist(out$p.re.level.names))),
sort(unique(c(X.p.re))))
})
# Check output data output is correct -
test_that("out$y == y", {
expect_equal(out$y, y)
})
# Check missing values ----------------
test_that("missing value error handling works", {
# tmp.data <- data.list
# tmp.data$occ.covs[3, ] <- NA
# expect_error(spMsPGOcc(occ.formula = occ.formula,
# det.formula = det.formula,
# data = tmp.data,
# n.batch = 40,
# batch.length = batch.length,
# cov.model = "exponential",
# tuning = tuning.list,
# NNGP = TRUE,
# verbose = FALSE,
# n.neighbors = 5,
# search.type = 'cb',
# n.report = 10,
# n.burn = 100,
# n.chains = 1))
tmp.data <- data.list
tmp.data$y[1, 1, 1] <- 1
tmp.data$det.covs[[1]][1] <- NA
expect_error(spMsPGOcc(occ.formula = occ.formula,
det.formula = det.formula,
data = tmp.data,
n.batch = 40,
batch.length = batch.length,
cov.model = "exponential",
tuning = tuning.list,
NNGP = TRUE,
verbose = FALSE,
n.neighbors = 5,
search.type = 'cb',
n.report = 10,
n.burn = 100,
n.chains = 1))
tmp.data <- data.list
tmp.data$y[, 1, 1] <- NA
out <- spMsPGOcc(occ.formula = occ.formula,
det.formula = det.formula,
data = tmp.data,
n.batch = 40,
batch.length = batch.length,
cov.model = "exponential",
tuning = tuning.list,
NNGP = TRUE,
verbose = FALSE,
n.neighbors = 5,
search.type = 'cb',
n.report = 10,
n.burn = 100,
n.chains = 1)
expect_s3_class(out, "spMsPGOcc")
})
# Check default values ----------------
test_that("default priors, inits, burn, thin work", {
out <- spMsPGOcc(occ.formula = occ.formula,
det.formula = det.formula,
data = data.list,
n.batch = n.batch,
batch.length = batch.length,
accept.rate = 0.43,
cov.model = "exponential",
tuning = tuning.list,
NNGP = TRUE,
verbose = FALSE,
n.neighbors = 5,
n.chains = 1)
expect_s3_class(out, "spMsPGOcc")
})
test_that("all correlation functions work", {
out <- spMsPGOcc(occ.formula = occ.formula,
det.formula = det.formula,
data = data.list,
inits = inits.list,
n.batch = n.batch,
batch.length = batch.length,
accept.rate = 0.43,
priors = prior.list,
cov.model = "gaussian",
tuning = list(phi = 0.3),
n.omp.threads = 1,
verbose = FALSE,
NNGP = TRUE,
n.neighbors = 5,
search.type = 'cb',
n.report = 10,
n.burn = 100,
n.thin = 1,
n.chains = 1)
expect_s3_class(out, "spMsPGOcc")
out <- spMsPGOcc(occ.formula = occ.formula,
det.formula = det.formula,
data = data.list,
inits = inits.list,
n.batch = n.batch,
batch.length = batch.length,
accept.rate = 0.43,
priors = prior.list,
cov.model = "spherical",
tuning = list(phi = 0.3),
n.omp.threads = 1,
verbose = FALSE,
NNGP = TRUE,
n.neighbors = 5,
search.type = 'cb',
n.report = 10,
n.burn = 100,
n.thin = 1,
n.chains = 1)
expect_s3_class(out, "spMsPGOcc")
out <- spMsPGOcc(occ.formula = occ.formula,
det.formula = det.formula,
data = data.list,
inits = inits.list,
n.batch = n.batch,
batch.length = batch.length,
accept.rate = 0.43,
priors = list(nu.unif = list(0.4, 3)),
cov.model = "matern",
tuning = list(phi = 0.3, nu = 0.2),
n.omp.threads = 1,
verbose = FALSE,
NNGP = TRUE,
n.neighbors = 5,
search.type = 'cb',
n.report = 10,
n.burn = 100,
n.thin = 1,
n.chains = 1)
expect_s3_class(out, "spMsPGOcc")
})
test_that("verbose prints to the screen", {
expect_output(spMsPGOcc(occ.formula = occ.formula,
det.formula = det.formula,
data = data.list,
inits = inits.list,
n.batch = n.batch,
batch.length = batch.length,
accept.rate = 0.43,
priors = prior.list,
cov.model = "exponential",
tuning = tuning.list,
n.omp.threads = 1,
verbose = TRUE,
NNGP = TRUE,
n.neighbors = 5,
search.type = 'cb',
n.report = 10,
n.burn = 100,
n.thin = 1,
n.chains = 1))
})
# Check waicOcc -----------------------
test_that("waicOCC works for spMsPGOcc", {
# as.vector gets rid of names
waic.out <- as.vector(waicOcc(out))
expect_equal(length(waic.out), 3)
expect_equal(waic.out[3], -2 * (waic.out[1] - waic.out[2]))
})
test_that("fitted works for spMsPGOcc", {
fitted.out <- fitted(out)
expect_equal(length(fitted.out), 2)
expect_equal(class(fitted.out$y.rep.samples), "array")
expect_equal(class(fitted.out$p.samples), "array")
expect_equal(dim(fitted.out$y.rep.samples), dim(fitted.out$p.samples))
})
test_that("predict works for spMsPGOcc", {
n.post.samples <- out$n.post * out$n.chains
X.0 <- cbind(X.0)
colnames(X.0) <- c('int')
pred.out <- predict(out, X.0, coords.0, verbose = FALSE)
expect_type(pred.out, "list")
expect_equal(dim(pred.out$psi.0.samples), c(n.post.samples, N, nrow(X.0)))
expect_equal(dim(pred.out$z.0.samples), c(n.post.samples, N, nrow(X.0)))
})
test_that("detection prediction works", {
X.p.0 <- cbind(dat$X.p[, 1, 1], dat$X.p.re[, 1, 1])
colnames(X.p.0) <- c('intercept', 'det.factor.1')
pred.out <- predict(out, X.p.0, type = 'detection')
expect_type(pred.out, 'list')
expect_equal(dim(pred.out$p.0.samples), c(out$n.post * out$n.chains, N, J))
})
test_that("posterior predictive checks work for msPGOcc", {
n.post.samples <- out$n.post * out$n.chains
J.fit <- nrow(X)
ppc.out <- ppcOcc(out, 'chi-square', 2)
expect_type(ppc.out, "list")
expect_equal(dim(ppc.out$fit.y), c(n.post.samples, N))
expect_equal(dim(ppc.out$fit.y.rep), c(n.post.samples, N))
expect_equal(dim(ppc.out$fit.y.group.quants), c(5, N, max(n.rep)))
expect_equal(dim(ppc.out$fit.y.rep.group.quants), c(5, N, max(n.rep)))
ppc.out <- ppcOcc(out, 'chi-square', 1)
expect_type(ppc.out, "list")
expect_equal(dim(ppc.out$fit.y), c(n.post.samples, N))
expect_equal(dim(ppc.out$fit.y.rep), c(n.post.samples, N))
expect_equal(dim(ppc.out$fit.y.group.quants), c(5, N, J.fit))
expect_equal(dim(ppc.out$fit.y.rep.group.quants), c(5, N, J.fit))
ppc.out <- ppcOcc(out, 'freeman-tukey', 1)
expect_type(ppc.out, "list")
expect_equal(dim(ppc.out$fit.y), c(n.post.samples, N))
expect_equal(dim(ppc.out$fit.y.rep), c(n.post.samples, N))
expect_equal(dim(ppc.out$fit.y.group.quants), c(5, N, J.fit))
expect_equal(dim(ppc.out$fit.y.rep.group.quants), c(5, N, J.fit))
ppc.out <- ppcOcc(out, 'freeman-tukey', 2)
expect_type(ppc.out, "list")
expect_equal(dim(ppc.out$fit.y), c(n.post.samples, N))
expect_equal(dim(ppc.out$fit.y.rep), c(n.post.samples, N))
expect_equal(dim(ppc.out$fit.y.group.quants), c(5, N, max(n.rep)))
expect_equal(dim(ppc.out$fit.y.rep.group.quants), c(5, N, max(n.rep)))
})
# Multiple random intercepts on detection ---------------------------------
J.x <- 8
J.y <- 8
J <- J.x * J.y
n.rep<- sample(2:4, size = J, replace = TRUE)
N <- 6
# Community-level covariate effects
# Occurrence
beta.mean <- c(0.2)
p.occ <- length(beta.mean)
tau.sq.beta <- c(0.6)
# Detection
alpha.mean <- c(0)
tau.sq.alpha <- c(1)
p.det <- length(alpha.mean)
# Random effects
psi.RE <- list()
p.RE <- list(levels = c(50, 10),
sigma.sq.p = c(2.50, 1.5))
# Draw species-level effects from community means.
beta <- matrix(NA, nrow = N, ncol = p.occ)
alpha <- matrix(NA, nrow = N, ncol = p.det)
for (i in 1:p.occ) {
beta[, i] <- rnorm(N, beta.mean[i], sqrt(tau.sq.beta[i]))
}
for (i in 1:p.det) {
alpha[, i] <- rnorm(N, alpha.mean[i], sqrt(tau.sq.alpha[i]))
}
alpha.true <- alpha
phi <- rep(3 / .7, N)
sigma.sq <- rep(2, N)
nu <- rep(2, N)
dat <- simMsOcc(J.x = J.x, J.y = J.y, n.rep = n.rep, N = N, beta = beta, alpha = alpha,
psi.RE = psi.RE, p.RE = p.RE, sp = TRUE, sigma.sq = sigma.sq,
phi = phi, nu = nu, cov.model = 'matern')
pred.indx <- sample(1:J, round(J * .25), replace = FALSE)
y <- dat$y[, -pred.indx, , drop = FALSE]
# Occupancy covariates
X <- dat$X[-pred.indx, , drop = FALSE]
# X.re <- dat$X.re[-pred.indx, , drop = FALSE]
coords <- as.matrix(dat$coords[-pred.indx, , drop = FALSE])
# Prediction covariates
X.0 <- dat$X[pred.indx, , drop = FALSE]
# X.re.0 <- dat$X.re[pred.indx, , drop = FALSE]
coords.0 <- as.matrix(dat$coords[pred.indx, , drop = FALSE])
# Detection covariates
X.p <- dat$X.p[-pred.indx, , , drop = FALSE]
X.p.re <- dat$X.p.re[-pred.indx, , , drop = FALSE]
occ.covs <- cbind(X)
colnames(occ.covs) <- c('int')
det.covs <- list(det.factor.1 = X.p.re[, , 1],
det.factor.2 = X.p.re[, , 2])
data.list <- list(y = y, coords = coords, occ.covs = occ.covs, det.covs = det.covs)
# Priors
prior.list <- list(beta.comm.normal = list(mean = 0, var = 2.72),
alpha.comm.normal = list(mean = 0, var = 2.72),
tau.sq.beta.ig = list(a = 0.1, b = 0.1),
tau.sq.alpha.ig = list(a = 0.1, b = 0.1),
nu.unif = list(0.5, 2.5))
# Starting values
inits.list <- list(alpha.comm = 0,
beta.comm = 0,
beta = 0,
alpha = 0,
tau.sq.beta = 1,
tau.sq.alpha = 1,
z = apply(y, c(1, 2), max, na.rm = TRUE))
# Tuning
tuning.list <- list(phi = 1, nu = 0.25)
batch.length <- 25
n.batch <- 10
n.report <- 100
occ.formula <- ~ 1
det.formula <- ~ (1 | det.factor.1) + (1 | det.factor.2)
out <- spMsPGOcc(occ.formula = occ.formula,
det.formula = det.formula,
data = data.list,
inits = inits.list,
n.batch = n.batch,
batch.length = batch.length,
accept.rate = 0.43,
priors = prior.list,
cov.model = "matern",
tuning = tuning.list,
n.omp.threads = 1,
verbose = FALSE,
NNGP = TRUE,
n.neighbors = 5,
search.type = 'cb',
n.report = 10,
n.burn = 100,
n.thin = 2,
n.chains = 2,
k.fold = 2,
k.fold.threads = 1)
# To make sure it worked --------------
test_that("out is of class spMsPGOcc", {
expect_s3_class(out, "spMsPGOcc")
})
# Check cross-validation --------------
test_that("cross-validation works", {
expect_equal(length(out$k.fold.deviance), N)
expect_type(out$k.fold.deviance, "double")
expect_equal(sum(out$k.fold.deviance < 0), 0)
})
# Check random effects ----------------
test_that("random effects are correct", {
expect_equal(out$pRE, TRUE)
expect_equal(out$psiRE, FALSE)
})
# Check RE error ----------------------
test_that("random effect gives error when non-numeric", {
# data.list$occ.covs <- as.data.frame(data.list$occ.covs)
# data.list$occ.covs$occ.factor.1 <- factor(data.list$occ.covs$occ.factor.1)
data.list$det.covs$det.factor.1 <- factor(data.list$det.covs$det.factor.1)
expect_error(out <- spMsPGOcc(occ.formula = occ.formula,
det.formula = det.formula,
data = data.list,
inits = inits.list,
n.batch = n.batch,
batch.length = batch.length,
accept.rate = 0.43,
priors = prior.list,
cov.model = "matern",
tuning = tuning.list,
n.omp.threads = 1,
verbose = FALSE,
NNGP = TRUE,
n.neighbors = 5,
search.type = 'cb',
n.report = 10,
n.burn = 100,
n.thin = 2,
n.chains = 1))
# data.list$occ.covs$occ.factor.1 <- as.character(factor(data.list$occ.covs$occ.factor.1))
data.list$det.covs$det.factor.1 <- as.character(factor(data.list$det.covs$det.factor.1))
expect_error(out <- spMsPGOcc(occ.formula = occ.formula,
det.formula = det.formula,
data = data.list,
inits = inits.list,
n.batch = n.batch,
batch.length = batch.length,
accept.rate = 0.43,
priors = prior.list,
cov.model = "matern",
tuning = tuning.list,
n.omp.threads = 1,
verbose = FALSE,
NNGP = TRUE,
n.neighbors = 5,
search.type = 'cb',
n.report = 10,
n.burn = 100,
n.thin = 2,
n.chains = 1))
})
# Check RE levels ---------------------
test_that("random effect levels are correct", {
expect_equal(sort(unique(unlist(out$p.re.level.names))),
sort(unique(c(X.p.re))))
})
# Check output data output is correct -
test_that("out$y == y", {
expect_equal(out$y, y)
})
# Check missing values ----------------
test_that("missing value error handling works", {
# tmp.data <- data.list
# tmp.data$occ.covs[3, ] <- NA
# expect_error(spMsPGOcc(occ.formula = occ.formula,
# det.formula = det.formula,
# data = tmp.data,
# n.batch = 40,
# batch.length = batch.length,
# cov.model = "exponential",
# tuning = tuning.list,
# NNGP = TRUE,
# verbose = FALSE,
# n.neighbors = 5,
# search.type = 'cb',
# n.report = 10,
# n.burn = 100,
# n.chains = 1))
tmp.data <- data.list
tmp.data$y[1, 1, 1] <- 1
tmp.data$det.covs[[1]][1] <- NA
expect_error(spMsPGOcc(occ.formula = occ.formula,
det.formula = det.formula,
data = tmp.data,
n.batch = 40,
batch.length = batch.length,
cov.model = "exponential",
tuning = tuning.list,
NNGP = TRUE,
verbose = FALSE,
n.neighbors = 5,
search.type = 'cb',
n.report = 10,
n.burn = 100,
n.chains = 1))
tmp.data <- data.list
tmp.data$y[, 1, 1] <- NA
out <- spMsPGOcc(occ.formula = occ.formula,
det.formula = det.formula,
data = tmp.data,
n.batch = 40,
batch.length = batch.length,
cov.model = "exponential",
tuning = tuning.list,
NNGP = TRUE,
verbose = FALSE,
n.neighbors = 5,
search.type = 'cb',
n.report = 10,
n.burn = 100,
n.chains = 1)
expect_s3_class(out, "spMsPGOcc")
})
# Check default values ----------------
test_that("default priors, inits, burn, thin work", {
out <- spMsPGOcc(occ.formula = occ.formula,
det.formula = det.formula,
data = data.list,
n.batch = n.batch,
batch.length = batch.length,
accept.rate = 0.43,
cov.model = "exponential",
tuning = tuning.list,
NNGP = TRUE,
verbose = FALSE,
n.neighbors = 5,
n.chains = 1)
expect_s3_class(out, "spMsPGOcc")
})
test_that("all correlation functions work", {
out <- spMsPGOcc(occ.formula = occ.formula,
det.formula = det.formula,
data = data.list,
inits = inits.list,
n.batch = n.batch,
batch.length = batch.length,
accept.rate = 0.43,
priors = prior.list,
cov.model = "gaussian",
tuning = list(phi = 0.3),
n.omp.threads = 1,
verbose = FALSE,
NNGP = TRUE,
n.neighbors = 5,
search.type = 'cb',
n.report = 10,
n.burn = 100,
n.thin = 1,
n.chains = 1)
expect_s3_class(out, "spMsPGOcc")
out <- spMsPGOcc(occ.formula = occ.formula,
det.formula = det.formula,
data = data.list,
inits = inits.list,
n.batch = n.batch,
batch.length = batch.length,
accept.rate = 0.43,
priors = prior.list,
cov.model = "spherical",
tuning = list(phi = 0.3),
n.omp.threads = 1,
verbose = FALSE,
NNGP = TRUE,
n.neighbors = 5,
search.type = 'cb',
n.report = 10,
n.burn = 100,
n.thin = 1,
n.chains = 1)
expect_s3_class(out, "spMsPGOcc")
out <- spMsPGOcc(occ.formula = occ.formula,
det.formula = det.formula,
data = data.list,
inits = inits.list,
n.batch = n.batch,
batch.length = batch.length,
accept.rate = 0.43,
priors = list(nu.unif = list(0.4, 3)),
cov.model = "matern",
tuning = list(phi = 0.3, nu = 0.2),
n.omp.threads = 1,
verbose = FALSE,
NNGP = TRUE,
n.neighbors = 5,
search.type = 'cb',
n.report = 10,
n.burn = 100,
n.thin = 1,
n.chains = 1)
expect_s3_class(out, "spMsPGOcc")
})
test_that("verbose prints to the screen", {
expect_output(spMsPGOcc(occ.formula = occ.formula,
det.formula = det.formula,
data = data.list,
inits = inits.list,
n.batch = n.batch,
batch.length = batch.length,
accept.rate = 0.43,
priors = prior.list,
cov.model = "exponential",
tuning = tuning.list,
n.omp.threads = 1,
verbose = TRUE,
NNGP = TRUE,
n.neighbors = 5,
search.type = 'cb',
n.report = 10,
n.burn = 100,
n.thin = 1,
n.chains = 1))
})
# Check waicOcc -----------------------
test_that("waicOCC works for spMsPGOcc", {
# as.vector gets rid of names
waic.out <- as.vector(waicOcc(out))
expect_equal(length(waic.out), 3)
expect_equal(waic.out[3], -2 * (waic.out[1] - waic.out[2]))
})
test_that("fitted works for spMsPGOcc", {
fitted.out <- fitted(out)
expect_equal(length(fitted.out), 2)
expect_equal(class(fitted.out$y.rep.samples), "array")
expect_equal(class(fitted.out$p.samples), "array")
expect_equal(dim(fitted.out$y.rep.samples), dim(fitted.out$p.samples))
})
test_that("predict works for spMsPGOcc", {
n.post.samples <- out$n.post * out$n.chains
X.0 <- cbind(X.0)
colnames(X.0) <- c('int')
pred.out <- predict(out, X.0, coords.0, verbose = FALSE)
expect_type(pred.out, "list")
expect_equal(dim(pred.out$psi.0.samples), c(n.post.samples, N, nrow(X.0)))
expect_equal(dim(pred.out$z.0.samples), c(n.post.samples, N, nrow(X.0)))
})
test_that("detection prediction works", {
X.p.0 <- cbind(dat$X.p[, 1, 1], dat$X.p.re[, 1, 1:2])
colnames(X.p.0) <- c('intercept', 'det.factor.1', 'det.factor.2')
pred.out <- predict(out, X.p.0, type = 'detection')
expect_type(pred.out, 'list')
expect_equal(dim(pred.out$p.0.samples), c(out$n.post * out$n.chains, N, J))
})
test_that("posterior predictive checks work for msPGOcc", {
n.post.samples <- out$n.post * out$n.chains
J.fit <- nrow(X)
ppc.out <- ppcOcc(out, 'chi-square', 2)
expect_type(ppc.out, "list")
expect_equal(dim(ppc.out$fit.y), c(n.post.samples, N))
expect_equal(dim(ppc.out$fit.y.rep), c(n.post.samples, N))
expect_equal(dim(ppc.out$fit.y.group.quants), c(5, N, max(n.rep)))
expect_equal(dim(ppc.out$fit.y.rep.group.quants), c(5, N, max(n.rep)))
ppc.out <- ppcOcc(out, 'chi-square', 1)
expect_type(ppc.out, "list")
expect_equal(dim(ppc.out$fit.y), c(n.post.samples, N))
expect_equal(dim(ppc.out$fit.y.rep), c(n.post.samples, N))
expect_equal(dim(ppc.out$fit.y.group.quants), c(5, N, J.fit))
expect_equal(dim(ppc.out$fit.y.rep.group.quants), c(5, N, J.fit))
ppc.out <- ppcOcc(out, 'freeman-tukey', 1)
expect_type(ppc.out, "list")
expect_equal(dim(ppc.out$fit.y), c(n.post.samples, N))
expect_equal(dim(ppc.out$fit.y.rep), c(n.post.samples, N))
expect_equal(dim(ppc.out$fit.y.group.quants), c(5, N, J.fit))
expect_equal(dim(ppc.out$fit.y.rep.group.quants), c(5, N, J.fit))
ppc.out <- ppcOcc(out, 'freeman-tukey', 2)
expect_type(ppc.out, "list")
expect_equal(dim(ppc.out$fit.y), c(n.post.samples, N))
expect_equal(dim(ppc.out$fit.y.rep), c(n.post.samples, N))
expect_equal(dim(ppc.out$fit.y.group.quants), c(5, N, max(n.rep)))
expect_equal(dim(ppc.out$fit.y.rep.group.quants), c(5, N, max(n.rep)))
})
# Detection random effects with covariates --------------------------------
J.x <- 8
J.y <- 8
J <- J.x * J.y
n.rep<- sample(2:4, size = J, replace = TRUE)
N <- 6
# Community-level covariate effects
# Occurrence
beta.mean <- c(0.2)
p.occ <- length(beta.mean)
tau.sq.beta <- c(0.6)
# Detection
alpha.mean <- c(0, 0.5)
tau.sq.alpha <- c(1, 2.3)
p.det <- length(alpha.mean)
# Random effects
psi.RE <- list()
p.RE <- list(levels = c(50, 10),
sigma.sq.p = c(2.50, 1.5))
# Draw species-level effects from community means.
beta <- matrix(NA, nrow = N, ncol = p.occ)
alpha <- matrix(NA, nrow = N, ncol = p.det)
for (i in 1:p.occ) {
beta[, i] <- rnorm(N, beta.mean[i], sqrt(tau.sq.beta[i]))
}
for (i in 1:p.det) {
alpha[, i] <- rnorm(N, alpha.mean[i], sqrt(tau.sq.alpha[i]))
}
alpha.true <- alpha
phi <- rep(3 / .7, N)
sigma.sq <- rep(2, N)
nu <- rep(2, N)
dat <- simMsOcc(J.x = J.x, J.y = J.y, n.rep = n.rep, N = N, beta = beta, alpha = alpha,
psi.RE = psi.RE, p.RE = p.RE, sp = TRUE, sigma.sq = sigma.sq,
phi = phi, nu = nu, cov.model = 'matern')
pred.indx <- sample(1:J, round(J * .25), replace = FALSE)
y <- dat$y[, -pred.indx, , drop = FALSE]
# Occupancy covariates
X <- dat$X[-pred.indx, , drop = FALSE]
# X.re <- dat$X.re[-pred.indx, , drop = FALSE]
coords <- as.matrix(dat$coords[-pred.indx, , drop = FALSE])
# Prediction covariates
X.0 <- dat$X[pred.indx, , drop = FALSE]
# X.re.0 <- dat$X.re[pred.indx, , drop = FALSE]
coords.0 <- as.matrix(dat$coords[pred.indx, , drop = FALSE])
# Detection covariates
X.p <- dat$X.p[-pred.indx, , , drop = FALSE]
X.p.re <- dat$X.p.re[-pred.indx, , , drop = FALSE]
occ.covs <- cbind(X)
colnames(occ.covs) <- c('int')
det.covs <- list(det.cov.1 = X.p[, , 2],
det.factor.1 = X.p.re[, , 1],
det.factor.2 = X.p.re[, , 2])
data.list <- list(y = y, coords = coords, occ.covs = occ.covs, det.covs = det.covs)
# Priors
prior.list <- list(beta.comm.normal = list(mean = 0, var = 2.72),
alpha.comm.normal = list(mean = 0, var = 2.72),
tau.sq.beta.ig = list(a = 0.1, b = 0.1),
tau.sq.alpha.ig = list(a = 0.1, b = 0.1),
nu.unif = list(0.5, 2.5))
# Starting values
inits.list <- list(alpha.comm = 0,
beta.comm = 0,
beta = 0,
alpha = 0,
tau.sq.beta = 1,
tau.sq.alpha = 1,
z = apply(y, c(1, 2), max, na.rm = TRUE))
# Tuning
tuning.list <- list(phi = 1, nu = 0.25)
batch.length <- 25
n.batch <- 10
n.report <- 100
occ.formula <- ~ 1
det.formula <- ~ det.cov.1 + (1 | det.factor.1) + (1 | det.factor.2)
out <- spMsPGOcc(occ.formula = occ.formula,
det.formula = det.formula,
data = data.list,
inits = inits.list,
n.batch = n.batch,
batch.length = batch.length,
accept.rate = 0.43,
priors = prior.list,
cov.model = "matern",
tuning = tuning.list,
n.omp.threads = 1,
verbose = FALSE,
NNGP = TRUE,
n.neighbors = 5,
search.type = 'cb',
n.report = 10,
n.burn = 100,
n.thin = 2,
n.chains = 2,
k.fold = 2,
k.fold.threads = 1)
# To make sure it worked --------------
test_that("out is of class spMsPGOcc", {
expect_s3_class(out, "spMsPGOcc")
})
# Check cross-validation --------------
test_that("cross-validation works", {
expect_equal(length(out$k.fold.deviance), N)
expect_type(out$k.fold.deviance, "double")
expect_equal(sum(out$k.fold.deviance < 0), 0)
})
# Check random effects ----------------
test_that("random effects are correct", {
expect_equal(out$pRE, TRUE)
expect_equal(out$psiRE, FALSE)
})
# Check RE error ----------------------
test_that("random effect gives error when non-numeric", {
# data.list$occ.covs <- as.data.frame(data.list$occ.covs)
# data.list$occ.covs$occ.factor.1 <- factor(data.list$occ.covs$occ.factor.1)
data.list$det.covs$det.factor.1 <- factor(data.list$det.covs$det.factor.1)
expect_error(out <- spMsPGOcc(occ.formula = occ.formula,
det.formula = det.formula,
data = data.list,
inits = inits.list,
n.batch = n.batch,
batch.length = batch.length,
accept.rate = 0.43,
priors = prior.list,
cov.model = "matern",
tuning = tuning.list,
n.omp.threads = 1,
verbose = FALSE,
NNGP = TRUE,
n.neighbors = 5,
search.type = 'cb',
n.report = 10,
n.burn = 100,
n.thin = 2,
n.chains = 1))
# data.list$occ.covs$occ.factor.1 <- as.character(factor(data.list$occ.covs$occ.factor.1))
data.list$det.covs$det.factor.1 <- as.character(factor(data.list$det.covs$det.factor.1))
expect_error(out <- spMsPGOcc(occ.formula = occ.formula,
det.formula = det.formula,
data = data.list,
inits = inits.list,
n.batch = n.batch,
batch.length = batch.length,
accept.rate = 0.43,
priors = prior.list,
cov.model = "matern",
tuning = tuning.list,
n.omp.threads = 1,
verbose = FALSE,
NNGP = TRUE,
n.neighbors = 5,
search.type = 'cb',
n.report = 10,
n.burn = 100,
n.thin = 2,
n.chains = 1))
})
# Check RE levels ---------------------
test_that("random effect levels are correct", {
expect_equal(sort(unique(unlist(out$p.re.level.names))),
sort(unique(c(X.p.re))))
})
# Check output data output is correct -
test_that("out$y == y", {
expect_equal(out$y, y)
})
# Check missing values ----------------
test_that("missing value error handling works", {
# tmp.data <- data.list
# tmp.data$occ.covs[3, ] <- NA
# expect_error(spMsPGOcc(occ.formula = occ.formula,
# det.formula = det.formula,
# data = tmp.data,
# n.batch = 40,
# batch.length = batch.length,
# cov.model = "exponential",
# tuning = tuning.list,
# NNGP = TRUE,
# verbose = FALSE,
# n.neighbors = 5,
# search.type = 'cb',
# n.report = 10,
# n.burn = 100,
# n.chains = 1))
tmp.data <- data.list
tmp.data$y[1, 1, 1] <- 1
tmp.data$det.covs[[1]][1] <- NA
expect_error(spMsPGOcc(occ.formula = occ.formula,
det.formula = det.formula,
data = tmp.data,
n.batch = 40,
batch.length = batch.length,
cov.model = "exponential",
tuning = tuning.list,
NNGP = TRUE,
verbose = FALSE,
n.neighbors = 5,
search.type = 'cb',
n.report = 10,
n.burn = 100,
n.chains = 1))
tmp.data <- data.list
tmp.data$y[, 1, 1] <- NA
out <- spMsPGOcc(occ.formula = occ.formula,
det.formula = det.formula,
data = tmp.data,
n.batch = 40,
batch.length = batch.length,
cov.model = "exponential",
tuning = tuning.list,
NNGP = TRUE,
verbose = FALSE,
n.neighbors = 5,
search.type = 'cb',
n.report = 10,
n.burn = 100,
n.chains = 1)
expect_s3_class(out, "spMsPGOcc")
})
# Check default values ----------------
test_that("default priors, inits, burn, thin work", {
out <- spMsPGOcc(occ.formula = occ.formula,
det.formula = det.formula,
data = data.list,
n.batch = n.batch,
batch.length = batch.length,
accept.rate = 0.43,
cov.model = "exponential",
tuning = tuning.list,
NNGP = TRUE,
verbose = FALSE,
n.neighbors = 5,
n.chains = 1)
expect_s3_class(out, "spMsPGOcc")
})
test_that("all correlation functions work", {
out <- spMsPGOcc(occ.formula = occ.formula,
det.formula = det.formula,
data = data.list,
inits = inits.list,
n.batch = n.batch,
batch.length = batch.length,
accept.rate = 0.43,
priors = prior.list,
cov.model = "gaussian",
tuning = list(phi = 0.3),
n.omp.threads = 1,
verbose = FALSE,
NNGP = TRUE,
n.neighbors = 5,
search.type = 'cb',
n.report = 10,
n.burn = 100,
n.thin = 1,
n.chains = 1)
expect_s3_class(out, "spMsPGOcc")
out <- spMsPGOcc(occ.formula = occ.formula,
det.formula = det.formula,
data = data.list,
inits = inits.list,
n.batch = n.batch,
batch.length = batch.length,
accept.rate = 0.43,
priors = prior.list,
cov.model = "spherical",
tuning = list(phi = 0.3),
n.omp.threads = 1,
verbose = FALSE,
NNGP = TRUE,
n.neighbors = 5,
search.type = 'cb',
n.report = 10,
n.burn = 100,
n.thin = 1,
n.chains = 1)
expect_s3_class(out, "spMsPGOcc")
out <- spMsPGOcc(occ.formula = occ.formula,
det.formula = det.formula,
data = data.list,
inits = inits.list,
n.batch = n.batch,
batch.length = batch.length,
accept.rate = 0.43,
priors = list(nu.unif = list(0.4, 3)),
cov.model = "matern",
tuning = list(phi = 0.3, nu = 0.2),
n.omp.threads = 1,
verbose = FALSE,
NNGP = TRUE,
n.neighbors = 5,
search.type = 'cb',
n.report = 10,
n.burn = 100,
n.thin = 1,
n.chains = 1)
expect_s3_class(out, "spMsPGOcc")
})
test_that("verbose prints to the screen", {
expect_output(spMsPGOcc(occ.formula = occ.formula,
det.formula = det.formula,
data = data.list,
inits = inits.list,
n.batch = n.batch,
batch.length = batch.length,
accept.rate = 0.43,
priors = prior.list,
cov.model = "exponential",
tuning = tuning.list,
n.omp.threads = 1,
verbose = TRUE,
NNGP = TRUE,
n.neighbors = 5,
search.type = 'cb',
n.report = 10,
n.burn = 100,
n.thin = 1,
n.chains = 1))
})
# Check waicOcc -----------------------
test_that("waicOCC works for spMsPGOcc", {
# as.vector gets rid of names
waic.out <- as.vector(waicOcc(out))
expect_equal(length(waic.out), 3)
expect_equal(waic.out[3], -2 * (waic.out[1] - waic.out[2]))
})
test_that("fitted works for spMsPGOcc", {
fitted.out <- fitted(out)
expect_equal(length(fitted.out), 2)
expect_equal(class(fitted.out$y.rep.samples), "array")
expect_equal(class(fitted.out$p.samples), "array")
expect_equal(dim(fitted.out$y.rep.samples), dim(fitted.out$p.samples))
})
test_that("predict works for spMsPGOcc", {
n.post.samples <- out$n.post * out$n.chains
X.0 <- cbind(X.0)
colnames(X.0) <- c('int')
pred.out <- predict(out, X.0, coords.0, verbose = FALSE)
expect_type(pred.out, "list")
expect_equal(dim(pred.out$psi.0.samples), c(n.post.samples, N, nrow(X.0)))
expect_equal(dim(pred.out$z.0.samples), c(n.post.samples, N, nrow(X.0)))
})
test_that("detection prediction works", {
X.p.0 <- cbind(dat$X.p[, 1, ], dat$X.p.re[, 1, 1:2])
colnames(X.p.0) <- c('intercept', 'det.cov.1', 'det.factor.1', 'det.factor.2')
pred.out <- predict(out, X.p.0, type = 'detection')
expect_type(pred.out, 'list')
expect_equal(dim(pred.out$p.0.samples), c(out$n.post * out$n.chains, N, J))
})
test_that("posterior predictive checks work for msPGOcc", {
n.post.samples <- out$n.post * out$n.chains
J.fit <- nrow(X)
ppc.out <- ppcOcc(out, 'chi-square', 2)
expect_type(ppc.out, "list")
expect_equal(dim(ppc.out$fit.y), c(n.post.samples, N))
expect_equal(dim(ppc.out$fit.y.rep), c(n.post.samples, N))
expect_equal(dim(ppc.out$fit.y.group.quants), c(5, N, max(n.rep)))
expect_equal(dim(ppc.out$fit.y.rep.group.quants), c(5, N, max(n.rep)))
ppc.out <- ppcOcc(out, 'chi-square', 1)
expect_type(ppc.out, "list")
expect_equal(dim(ppc.out$fit.y), c(n.post.samples, N))
expect_equal(dim(ppc.out$fit.y.rep), c(n.post.samples, N))
expect_equal(dim(ppc.out$fit.y.group.quants), c(5, N, J.fit))
expect_equal(dim(ppc.out$fit.y.rep.group.quants), c(5, N, J.fit))
ppc.out <- ppcOcc(out, 'freeman-tukey', 1)
expect_type(ppc.out, "list")
expect_equal(dim(ppc.out$fit.y), c(n.post.samples, N))
expect_equal(dim(ppc.out$fit.y.rep), c(n.post.samples, N))
expect_equal(dim(ppc.out$fit.y.group.quants), c(5, N, J.fit))
expect_equal(dim(ppc.out$fit.y.rep.group.quants), c(5, N, J.fit))
ppc.out <- ppcOcc(out, 'freeman-tukey', 2)
expect_type(ppc.out, "list")
expect_equal(dim(ppc.out$fit.y), c(n.post.samples, N))
expect_equal(dim(ppc.out$fit.y.rep), c(n.post.samples, N))
expect_equal(dim(ppc.out$fit.y.group.quants), c(5, N, max(n.rep)))
expect_equal(dim(ppc.out$fit.y.rep.group.quants), c(5, N, max(n.rep)))
})
# Detection random effects with covariates on all -------------------------
J.x <- 8
J.y <- 8
J <- J.x * J.y
n.rep<- sample(2:4, size = J, replace = TRUE)
N <- 6
# Community-level covariate effects
# Occurrence
beta.mean <- c(0.2, 1.5, 0.3)
p.occ <- length(beta.mean)
tau.sq.beta <- c(0.6, 0.5, 3.3)
# Detection
alpha.mean <- c(0, 0.5)
tau.sq.alpha <- c(1, 2.3)
p.det <- length(alpha.mean)
# Random effects
psi.RE <- list()
p.RE <- list(levels = c(50, 10),
sigma.sq.p = c(2.50, 1.5))
# Draw species-level effects from community means.
beta <- matrix(NA, nrow = N, ncol = p.occ)
alpha <- matrix(NA, nrow = N, ncol = p.det)
for (i in 1:p.occ) {
beta[, i] <- rnorm(N, beta.mean[i], sqrt(tau.sq.beta[i]))
}
for (i in 1:p.det) {
alpha[, i] <- rnorm(N, alpha.mean[i], sqrt(tau.sq.alpha[i]))
}
alpha.true <- alpha
phi <- rep(3 / .7, N)
sigma.sq <- rep(2, N)
nu <- rep(2, N)
dat <- simMsOcc(J.x = J.x, J.y = J.y, n.rep = n.rep, N = N, beta = beta, alpha = alpha,
psi.RE = psi.RE, p.RE = p.RE, sp = TRUE, sigma.sq = sigma.sq,
phi = phi, nu = nu, cov.model = 'matern')
pred.indx <- sample(1:J, round(J * .25), replace = FALSE)
y <- dat$y[, -pred.indx, , drop = FALSE]
# Occupancy covariates
X <- dat$X[-pred.indx, , drop = FALSE]
# X.re <- dat$X.re[-pred.indx, , drop = FALSE]
coords <- as.matrix(dat$coords[-pred.indx, , drop = FALSE])
# Prediction covariates
X.0 <- dat$X[pred.indx, , drop = FALSE]
# X.re.0 <- dat$X.re[pred.indx, , drop = FALSE]
coords.0 <- as.matrix(dat$coords[pred.indx, , drop = FALSE])
# Detection covariates
X.p <- dat$X.p[-pred.indx, , , drop = FALSE]
X.p.re <- dat$X.p.re[-pred.indx, , , drop = FALSE]
occ.covs <- cbind(X)
colnames(occ.covs) <- c('int', 'occ.cov.1', 'occ.cov.2')
det.covs <- list(det.cov.1 = X.p[, , 2],
det.factor.1 = X.p.re[, , 1],
det.factor.2 = X.p.re[, , 2])
data.list <- list(y = y, coords = coords, occ.covs = occ.covs, det.covs = det.covs)
# Priors
prior.list <- list(beta.comm.normal = list(mean = 0, var = 2.72),
alpha.comm.normal = list(mean = 0, var = 2.72),
tau.sq.beta.ig = list(a = 0.1, b = 0.1),
tau.sq.alpha.ig = list(a = 0.1, b = 0.1),
nu.unif = list(0.5, 2.5))
# Starting values
inits.list <- list(alpha.comm = 0,
beta.comm = 0,
beta = 0,
alpha = 0,
tau.sq.beta = 1,
tau.sq.alpha = 1,
z = apply(y, c(1, 2), max, na.rm = TRUE))
# Tuning
tuning.list <- list(phi = 1, nu = 0.25)
batch.length <- 25
n.batch <- 10
n.report <- 100
occ.formula <- ~ occ.cov.1 + occ.cov.2
det.formula <- ~ det.cov.1 + (1 | det.factor.1) + (1 | det.factor.2)
out <- spMsPGOcc(occ.formula = occ.formula,
det.formula = det.formula,
data = data.list,
inits = inits.list,
n.batch = n.batch,
batch.length = batch.length,
accept.rate = 0.43,
priors = prior.list,
cov.model = "matern",
tuning = tuning.list,
n.omp.threads = 1,
verbose = FALSE,
NNGP = TRUE,
n.neighbors = 5,
search.type = 'cb',
n.report = 10,
n.burn = 100,
n.thin = 2,
n.chains = 2,
k.fold = 2,
k.fold.threads = 1)
# To make sure it worked --------------
test_that("out is of class spMsPGOcc", {
expect_s3_class(out, "spMsPGOcc")
})
# Check cross-validation --------------
test_that("cross-validation works", {
expect_equal(length(out$k.fold.deviance), N)
expect_type(out$k.fold.deviance, "double")
expect_equal(sum(out$k.fold.deviance < 0), 0)
})
# Check random effects ----------------
test_that("random effects are correct", {
expect_equal(out$pRE, TRUE)
expect_equal(out$psiRE, FALSE)
})
# Check RE error ----------------------
test_that("random effect gives error when non-numeric", {
# data.list$occ.covs <- as.data.frame(data.list$occ.covs)
# data.list$occ.covs$occ.factor.1 <- factor(data.list$occ.covs$occ.factor.1)
data.list$det.covs$det.factor.1 <- factor(data.list$det.covs$det.factor.1)
expect_error(out <- spMsPGOcc(occ.formula = occ.formula,
det.formula = det.formula,
data = data.list,
inits = inits.list,
n.batch = n.batch,
batch.length = batch.length,
accept.rate = 0.43,
priors = prior.list,
cov.model = "matern",
tuning = tuning.list,
n.omp.threads = 1,
verbose = FALSE,
NNGP = TRUE,
n.neighbors = 5,
search.type = 'cb',
n.report = 10,
n.burn = 100,
n.thin = 2,
n.chains = 1))
# data.list$occ.covs$occ.factor.1 <- as.character(factor(data.list$occ.covs$occ.factor.1))
data.list$det.covs$det.factor.1 <- as.character(factor(data.list$det.covs$det.factor.1))
expect_error(out <- spMsPGOcc(occ.formula = occ.formula,
det.formula = det.formula,
data = data.list,
inits = inits.list,
n.batch = n.batch,
batch.length = batch.length,
accept.rate = 0.43,
priors = prior.list,
cov.model = "matern",
tuning = tuning.list,
n.omp.threads = 1,
verbose = FALSE,
NNGP = TRUE,
n.neighbors = 5,
search.type = 'cb',
n.report = 10,
n.burn = 100,
n.thin = 2,
n.chains = 1))
})
# Check RE levels ---------------------
test_that("random effect levels are correct", {
expect_equal(sort(unique(unlist(out$p.re.level.names))),
sort(unique(c(X.p.re))))
})
# Check output data output is correct -
test_that("out$y == y", {
expect_equal(out$y, y)
})
# Check missing values ----------------
test_that("missing value error handling works", {
tmp.data <- data.list
tmp.data$occ.covs[3, ] <- NA
expect_error(spMsPGOcc(occ.formula = occ.formula,
det.formula = det.formula,
data = tmp.data,
n.batch = 40,
batch.length = batch.length,
cov.model = "exponential",
tuning = tuning.list,
NNGP = TRUE,
verbose = FALSE,
n.neighbors = 5,
search.type = 'cb',
n.report = 10,
n.burn = 100,
n.chains = 1))
tmp.data <- data.list
tmp.data$y[1, 1, 1] <- 1
tmp.data$det.covs[[1]][1] <- NA
expect_error(spMsPGOcc(occ.formula = occ.formula,
det.formula = det.formula,
data = tmp.data,
n.batch = 40,
batch.length = batch.length,
cov.model = "exponential",
tuning = tuning.list,
NNGP = TRUE,
verbose = FALSE,
n.neighbors = 5,
search.type = 'cb',
n.report = 10,
n.burn = 100,
n.chains = 1))
tmp.data <- data.list
tmp.data$y[, 1, 1] <- NA
out <- spMsPGOcc(occ.formula = occ.formula,
det.formula = det.formula,
data = tmp.data,
n.batch = 40,
batch.length = batch.length,
cov.model = "exponential",
tuning = tuning.list,
NNGP = TRUE,
verbose = FALSE,
n.neighbors = 5,
search.type = 'cb',
n.report = 10,
n.burn = 100,
n.chains = 1)
expect_s3_class(out, "spMsPGOcc")
})
# Check default values ----------------
test_that("default priors, inits, burn, thin work", {
out <- spMsPGOcc(occ.formula = occ.formula,
det.formula = det.formula,
data = data.list,
n.batch = n.batch,
batch.length = batch.length,
accept.rate = 0.43,
cov.model = "exponential",
tuning = tuning.list,
NNGP = TRUE,
verbose = FALSE,
n.neighbors = 5,
n.chains = 1)
expect_s3_class(out, "spMsPGOcc")
})
test_that("all correlation functions work", {
out <- spMsPGOcc(occ.formula = occ.formula,
det.formula = det.formula,
data = data.list,
inits = inits.list,
n.batch = n.batch,
batch.length = batch.length,
accept.rate = 0.43,
priors = prior.list,
cov.model = "gaussian",
tuning = list(phi = 0.3),
n.omp.threads = 1,
verbose = FALSE,
NNGP = TRUE,
n.neighbors = 5,
search.type = 'cb',
n.report = 10,
n.burn = 100,
n.thin = 1,
n.chains = 1)
expect_s3_class(out, "spMsPGOcc")
out <- spMsPGOcc(occ.formula = occ.formula,
det.formula = det.formula,
data = data.list,
inits = inits.list,
n.batch = n.batch,
batch.length = batch.length,
accept.rate = 0.43,
priors = prior.list,
cov.model = "spherical",
tuning = list(phi = 0.3),
n.omp.threads = 1,
verbose = FALSE,
NNGP = TRUE,
n.neighbors = 5,
search.type = 'cb',
n.report = 10,
n.burn = 100,
n.thin = 1,
n.chains = 1)
expect_s3_class(out, "spMsPGOcc")
out <- spMsPGOcc(occ.formula = occ.formula,
det.formula = det.formula,
data = data.list,
inits = inits.list,
n.batch = n.batch,
batch.length = batch.length,
accept.rate = 0.43,
priors = list(nu.unif = list(0.4, 3)),
cov.model = "matern",
tuning = list(phi = 0.3, nu = 0.2),
n.omp.threads = 1,
verbose = FALSE,
NNGP = TRUE,
n.neighbors = 5,
search.type = 'cb',
n.report = 10,
n.burn = 100,
n.thin = 1,
n.chains = 1)
expect_s3_class(out, "spMsPGOcc")
})
test_that("verbose prints to the screen", {
expect_output(spMsPGOcc(occ.formula = occ.formula,
det.formula = det.formula,
data = data.list,
inits = inits.list,
n.batch = n.batch,
batch.length = batch.length,
accept.rate = 0.43,
priors = prior.list,
cov.model = "exponential",
tuning = tuning.list,
n.omp.threads = 1,
verbose = TRUE,
NNGP = TRUE,
n.neighbors = 5,
search.type = 'cb',
n.report = 10,
n.burn = 100,
n.thin = 1,
n.chains = 1))
})
# Check waicOcc -----------------------
test_that("waicOCC works for spMsPGOcc", {
# as.vector gets rid of names
waic.out <- as.vector(waicOcc(out))
expect_equal(length(waic.out), 3)
expect_equal(waic.out[3], -2 * (waic.out[1] - waic.out[2]))
})
test_that("fitted works for spMsPGOcc", {
fitted.out <- fitted(out)
expect_equal(length(fitted.out), 2)
expect_equal(class(fitted.out$y.rep.samples), "array")
expect_equal(class(fitted.out$p.samples), "array")
expect_equal(dim(fitted.out$y.rep.samples), dim(fitted.out$p.samples))
})
test_that("predict works for spMsPGOcc", {
n.post.samples <- out$n.post * out$n.chains
X.0 <- cbind(X.0)
colnames(X.0) <- c('int', 'occ.cov.1', 'occ.cov.2')
pred.out <- predict(out, X.0, coords.0, verbose = FALSE)
expect_type(pred.out, "list")
expect_equal(dim(pred.out$psi.0.samples), c(n.post.samples, N, nrow(X.0)))
expect_equal(dim(pred.out$z.0.samples), c(n.post.samples, N, nrow(X.0)))
})
test_that("detection prediction works", {
X.p.0 <- cbind(dat$X.p[, 1, ], dat$X.p.re[, 1, 1:2])
colnames(X.p.0) <- c('intercept', 'det.cov.1', 'det.factor.1', 'det.factor.2')
pred.out <- predict(out, X.p.0, type = 'detection')
expect_type(pred.out, 'list')
expect_equal(dim(pred.out$p.0.samples), c(out$n.post * out$n.chains, N, J))
})
test_that("posterior predictive checks work for msPGOcc", {
n.post.samples <- out$n.post * out$n.chains
J.fit <- nrow(X)
ppc.out <- ppcOcc(out, 'chi-square', 2)
expect_type(ppc.out, "list")
expect_equal(dim(ppc.out$fit.y), c(n.post.samples, N))
expect_equal(dim(ppc.out$fit.y.rep), c(n.post.samples, N))
expect_equal(dim(ppc.out$fit.y.group.quants), c(5, N, max(n.rep)))
expect_equal(dim(ppc.out$fit.y.rep.group.quants), c(5, N, max(n.rep)))
ppc.out <- ppcOcc(out, 'chi-square', 1)
expect_type(ppc.out, "list")
expect_equal(dim(ppc.out$fit.y), c(n.post.samples, N))
expect_equal(dim(ppc.out$fit.y.rep), c(n.post.samples, N))
expect_equal(dim(ppc.out$fit.y.group.quants), c(5, N, J.fit))
expect_equal(dim(ppc.out$fit.y.rep.group.quants), c(5, N, J.fit))
ppc.out <- ppcOcc(out, 'freeman-tukey', 1)
expect_type(ppc.out, "list")
expect_equal(dim(ppc.out$fit.y), c(n.post.samples, N))
expect_equal(dim(ppc.out$fit.y.rep), c(n.post.samples, N))
expect_equal(dim(ppc.out$fit.y.group.quants), c(5, N, J.fit))
expect_equal(dim(ppc.out$fit.y.rep.group.quants), c(5, N, J.fit))
ppc.out <- ppcOcc(out, 'freeman-tukey', 2)
expect_type(ppc.out, "list")
expect_equal(dim(ppc.out$fit.y), c(n.post.samples, N))
expect_equal(dim(ppc.out$fit.y.rep), c(n.post.samples, N))
expect_equal(dim(ppc.out$fit.y.group.quants), c(5, N, max(n.rep)))
expect_equal(dim(ppc.out$fit.y.rep.group.quants), c(5, N, max(n.rep)))
})
# Random intercepts on both -----------------------------------------------
J.x <- 8
J.y <- 8
J <- J.x * J.y
n.rep<- sample(2:4, size = J, replace = TRUE)
N <- 6
# Community-level covariate effects
# Occurrence
beta.mean <- c(0.2)
p.occ <- length(beta.mean)
tau.sq.beta <- c(0.6)
# Detection
alpha.mean <- c(0)
tau.sq.alpha <- c(1)
p.det <- length(alpha.mean)
# Random effects
psi.RE <- list(levels = c(20),
sigma.sq.psi = c(2.5))
p.RE <- list(levels = c(50, 10),
sigma.sq.p = c(2.50, 1.5))
# Draw species-level effects from community means.
beta <- matrix(NA, nrow = N, ncol = p.occ)
alpha <- matrix(NA, nrow = N, ncol = p.det)
for (i in 1:p.occ) {
beta[, i] <- rnorm(N, beta.mean[i], sqrt(tau.sq.beta[i]))
}
for (i in 1:p.det) {
alpha[, i] <- rnorm(N, alpha.mean[i], sqrt(tau.sq.alpha[i]))
}
alpha.true <- alpha
phi <- rep(3 / .7, N)
sigma.sq <- rep(2, N)
nu <- rep(2, N)
dat <- simMsOcc(J.x = J.x, J.y = J.y, n.rep = n.rep, N = N, beta = beta, alpha = alpha,
psi.RE = psi.RE, p.RE = p.RE, sp = TRUE, sigma.sq = sigma.sq,
phi = phi, nu = nu, cov.model = 'matern')
pred.indx <- sample(1:J, round(J * .25), replace = FALSE)
y <- dat$y[, -pred.indx, , drop = FALSE]
# Occupancy covariates
X <- dat$X[-pred.indx, , drop = FALSE]
X.re <- dat$X.re[-pred.indx, , drop = FALSE]
coords <- as.matrix(dat$coords[-pred.indx, , drop = FALSE])
# Prediction covariates
X.0 <- dat$X[pred.indx, , drop = FALSE]
X.re.0 <- dat$X.re[pred.indx, , drop = FALSE]
coords.0 <- as.matrix(dat$coords[pred.indx, , drop = FALSE])
# Detection covariates
X.p <- dat$X.p[-pred.indx, , , drop = FALSE]
X.p.re <- dat$X.p.re[-pred.indx, , , drop = FALSE]
occ.covs <- cbind(X, X.re)
colnames(occ.covs) <- c('int', 'occ.factor.1')
det.covs <- list(det.factor.1 = X.p.re[, , 1],
det.factor.2 = X.p.re[, , 2])
data.list <- list(y = y, coords = coords, occ.covs = occ.covs, det.covs = det.covs)
# Priors
prior.list <- list(beta.comm.normal = list(mean = 0, var = 2.72),
alpha.comm.normal = list(mean = 0, var = 2.72),
tau.sq.beta.ig = list(a = 0.1, b = 0.1),
tau.sq.alpha.ig = list(a = 0.1, b = 0.1),
nu.unif = list(0.5, 2.5))
# Starting values
inits.list <- list(alpha.comm = 0,
beta.comm = 0,
beta = 0,
alpha = 0,
tau.sq.beta = 1,
tau.sq.alpha = 1,
z = apply(y, c(1, 2), max, na.rm = TRUE))
# Tuning
tuning.list <- list(phi = 1, nu = 0.25)
batch.length <- 25
n.batch <- 10
n.report <- 100
occ.formula <- ~ (1 | occ.factor.1)
det.formula <- ~ (1 | det.factor.1) + (1 | det.factor.2)
out <- spMsPGOcc(occ.formula = occ.formula,
det.formula = det.formula,
data = data.list,
inits = inits.list,
n.batch = n.batch,
batch.length = batch.length,
accept.rate = 0.43,
priors = prior.list,
cov.model = "matern",
tuning = tuning.list,
n.omp.threads = 1,
verbose = FALSE,
NNGP = TRUE,
n.neighbors = 5,
search.type = 'cb',
n.report = 10,
n.burn = 100,
n.thin = 2,
n.chains = 2,
k.fold = 2,
k.fold.threads = 1)
# To make sure it worked --------------
test_that("out is of class spMsPGOcc", {
expect_s3_class(out, "spMsPGOcc")
})
# Check cross-validation --------------
test_that("cross-validation works", {
expect_equal(length(out$k.fold.deviance), N)
expect_type(out$k.fold.deviance, "double")
expect_equal(sum(out$k.fold.deviance < 0), 0)
})
# Check random effects ----------------
test_that("random effects are correct", {
expect_equal(out$pRE, TRUE)
expect_equal(out$psiRE, TRUE)
})
# Check RE error ----------------------
test_that("random effect gives error when non-numeric", {
data.list$occ.covs <- as.data.frame(data.list$occ.covs)
data.list$occ.covs$occ.factor.1 <- factor(data.list$occ.covs$occ.factor.1)
data.list$det.covs$det.factor.1 <- factor(data.list$det.covs$det.factor.1)
expect_error(out <- spMsPGOcc(occ.formula = occ.formula,
det.formula = det.formula,
data = data.list,
inits = inits.list,
n.batch = n.batch,
batch.length = batch.length,
accept.rate = 0.43,
priors = prior.list,
cov.model = "matern",
tuning = tuning.list,
n.omp.threads = 1,
verbose = FALSE,
NNGP = TRUE,
n.neighbors = 5,
search.type = 'cb',
n.report = 10,
n.burn = 100,
n.thin = 2,
n.chains = 1))
data.list$occ.covs$occ.factor.1 <- as.character(factor(data.list$occ.covs$occ.factor.1))
data.list$det.covs$det.factor.1 <- as.character(factor(data.list$det.covs$det.factor.1))
expect_error(out <- spMsPGOcc(occ.formula = occ.formula,
det.formula = det.formula,
data = data.list,
inits = inits.list,
n.batch = n.batch,
batch.length = batch.length,
accept.rate = 0.43,
priors = prior.list,
cov.model = "matern",
tuning = tuning.list,
n.omp.threads = 1,
verbose = FALSE,
NNGP = TRUE,
n.neighbors = 5,
search.type = 'cb',
n.report = 10,
n.burn = 100,
n.thin = 2,
n.chains = 1))
})
# Check RE levels ---------------------
test_that("random effect levels are correct", {
expect_equal(sort(unique(unlist(out$re.level.names))),
sort(unique(c(X.re))))
expect_equal(sort(unique(unlist(out$p.re.level.names))),
sort(unique(c(X.p.re))))
})
# Check output data output is correct -
test_that("out$y == y", {
expect_equal(out$y, y)
})
# Check missing values ----------------
test_that("missing value error handling works", {
tmp.data <- data.list
tmp.data$occ.covs[3, ] <- NA
expect_error(spMsPGOcc(occ.formula = occ.formula,
det.formula = det.formula,
data = tmp.data,
n.batch = 40,
batch.length = batch.length,
cov.model = "exponential",
tuning = tuning.list,
NNGP = TRUE,
verbose = FALSE,
n.neighbors = 5,
search.type = 'cb',
n.report = 10,
n.burn = 100,
n.chains = 1))
tmp.data <- data.list
tmp.data$y[1, 1, 1] <- 1
tmp.data$det.covs[[1]][1] <- NA
expect_error(spMsPGOcc(occ.formula = occ.formula,
det.formula = det.formula,
data = tmp.data,
n.batch = 40,
batch.length = batch.length,
cov.model = "exponential",
tuning = tuning.list,
NNGP = TRUE,
verbose = FALSE,
n.neighbors = 5,
search.type = 'cb',
n.report = 10,
n.burn = 100,
n.chains = 1))
tmp.data <- data.list
tmp.data$y[, 1, 1] <- NA
out <- spMsPGOcc(occ.formula = occ.formula,
det.formula = det.formula,
data = tmp.data,
n.batch = 40,
batch.length = batch.length,
cov.model = "exponential",
tuning = tuning.list,
NNGP = TRUE,
verbose = FALSE,
n.neighbors = 5,
search.type = 'cb',
n.report = 10,
n.burn = 100,
n.chains = 1)
expect_s3_class(out, "spMsPGOcc")
})
# Check default values ----------------
test_that("default priors, inits, burn, thin work", {
out <- spMsPGOcc(occ.formula = occ.formula,
det.formula = det.formula,
data = data.list,
n.batch = n.batch,
batch.length = batch.length,
accept.rate = 0.43,
cov.model = "exponential",
tuning = tuning.list,
NNGP = TRUE,
verbose = FALSE,
n.neighbors = 5,
n.chains = 1)
expect_s3_class(out, "spMsPGOcc")
})
test_that("all correlation functions work", {
out <- spMsPGOcc(occ.formula = occ.formula,
det.formula = det.formula,
data = data.list,
inits = inits.list,
n.batch = n.batch,
batch.length = batch.length,
accept.rate = 0.43,
priors = prior.list,
cov.model = "gaussian",
tuning = list(phi = 0.3),
n.omp.threads = 1,
verbose = FALSE,
NNGP = TRUE,
n.neighbors = 5,
search.type = 'cb',
n.report = 10,
n.burn = 100,
n.thin = 1,
n.chains = 1)
expect_s3_class(out, "spMsPGOcc")
out <- spMsPGOcc(occ.formula = occ.formula,
det.formula = det.formula,
data = data.list,
inits = inits.list,
n.batch = n.batch,
batch.length = batch.length,
accept.rate = 0.43,
priors = prior.list,
cov.model = "spherical",
tuning = list(phi = 0.3),
n.omp.threads = 1,
verbose = FALSE,
NNGP = TRUE,
n.neighbors = 5,
search.type = 'cb',
n.report = 10,
n.burn = 100,
n.thin = 1,
n.chains = 1)
expect_s3_class(out, "spMsPGOcc")
out <- spMsPGOcc(occ.formula = occ.formula,
det.formula = det.formula,
data = data.list,
inits = inits.list,
n.batch = n.batch,
batch.length = batch.length,
accept.rate = 0.43,
priors = list(nu.unif = list(0.4, 3)),
cov.model = "matern",
tuning = list(phi = 0.3, nu = 0.2),
n.omp.threads = 1,
verbose = FALSE,
NNGP = TRUE,
n.neighbors = 5,
search.type = 'cb',
n.report = 10,
n.burn = 100,
n.thin = 1,
n.chains = 1)
expect_s3_class(out, "spMsPGOcc")
})
test_that("verbose prints to the screen", {
expect_output(spMsPGOcc(occ.formula = occ.formula,
det.formula = det.formula,
data = data.list,
inits = inits.list,
n.batch = n.batch,
batch.length = batch.length,
accept.rate = 0.43,
priors = prior.list,
cov.model = "exponential",
tuning = tuning.list,
n.omp.threads = 1,
verbose = TRUE,
NNGP = TRUE,
n.neighbors = 5,
search.type = 'cb',
n.report = 10,
n.burn = 100,
n.thin = 1,
n.chains = 1))
})
# Check waicOcc -----------------------
test_that("waicOCC works for spMsPGOcc", {
# as.vector gets rid of names
waic.out <- as.vector(waicOcc(out))
expect_equal(length(waic.out), 3)
expect_equal(waic.out[3], -2 * (waic.out[1] - waic.out[2]))
})
test_that("fitted works for spMsPGOcc", {
fitted.out <- fitted(out)
expect_equal(length(fitted.out), 2)
expect_equal(class(fitted.out$y.rep.samples), "array")
expect_equal(class(fitted.out$p.samples), "array")
expect_equal(dim(fitted.out$y.rep.samples), dim(fitted.out$p.samples))
})
test_that("predict works for spMsPGOcc", {
n.post.samples <- out$n.post * out$n.chains
X.0 <- cbind(X.0, X.re.0)
colnames(X.0) <- c('int', 'occ.factor.1')
pred.out <- predict(out, X.0, coords.0, verbose = FALSE)
expect_type(pred.out, "list")
expect_equal(dim(pred.out$psi.0.samples), c(n.post.samples, N, nrow(X.0)))
expect_equal(dim(pred.out$z.0.samples), c(n.post.samples, N, nrow(X.0)))
})
test_that("detection prediction works", {
X.p.0 <- cbind(dat$X.p[, 1, ], dat$X.p.re[, 1, 1:2])
colnames(X.p.0) <- c('intercept', 'det.factor.1', 'det.factor.2')
pred.out <- predict(out, X.p.0, type = 'detection')
expect_type(pred.out, 'list')
expect_equal(dim(pred.out$p.0.samples), c(out$n.post * out$n.chains, N, J))
})
test_that("posterior predictive checks work for msPGOcc", {
n.post.samples <- out$n.post * out$n.chains
J.fit <- nrow(X)
ppc.out <- ppcOcc(out, 'chi-square', 2)
expect_type(ppc.out, "list")
expect_equal(dim(ppc.out$fit.y), c(n.post.samples, N))
expect_equal(dim(ppc.out$fit.y.rep), c(n.post.samples, N))
expect_equal(dim(ppc.out$fit.y.group.quants), c(5, N, max(n.rep)))
expect_equal(dim(ppc.out$fit.y.rep.group.quants), c(5, N, max(n.rep)))
ppc.out <- ppcOcc(out, 'chi-square', 1)
expect_type(ppc.out, "list")
expect_equal(dim(ppc.out$fit.y), c(n.post.samples, N))
expect_equal(dim(ppc.out$fit.y.rep), c(n.post.samples, N))
expect_equal(dim(ppc.out$fit.y.group.quants), c(5, N, J.fit))
expect_equal(dim(ppc.out$fit.y.rep.group.quants), c(5, N, J.fit))
ppc.out <- ppcOcc(out, 'freeman-tukey', 1)
expect_type(ppc.out, "list")
expect_equal(dim(ppc.out$fit.y), c(n.post.samples, N))
expect_equal(dim(ppc.out$fit.y.rep), c(n.post.samples, N))
expect_equal(dim(ppc.out$fit.y.group.quants), c(5, N, J.fit))
expect_equal(dim(ppc.out$fit.y.rep.group.quants), c(5, N, J.fit))
ppc.out <- ppcOcc(out, 'freeman-tukey', 2)
expect_type(ppc.out, "list")
expect_equal(dim(ppc.out$fit.y), c(n.post.samples, N))
expect_equal(dim(ppc.out$fit.y.rep), c(n.post.samples, N))
expect_equal(dim(ppc.out$fit.y.group.quants), c(5, N, max(n.rep)))
expect_equal(dim(ppc.out$fit.y.rep.group.quants), c(5, N, max(n.rep)))
})
# Random intercepts on both with covariates -------------------------------
J.x <- 8
J.y <- 8
J <- J.x * J.y
n.rep<- sample(2:4, size = J, replace = TRUE)
N <- 6
# Community-level covariate effects
# Occurrence
beta.mean <- c(0.2, 1.2, -0.5)
p.occ <- length(beta.mean)
tau.sq.beta <- c(0.6, 2.4, 0.3)
# Detection
alpha.mean <- c(0, 0.5)
tau.sq.alpha <- c(1, 3.3)
p.det <- length(alpha.mean)
# Random effects
psi.RE <- list(levels = c(20),
sigma.sq.psi = c(2.5))
p.RE <- list(levels = c(50, 10),
sigma.sq.p = c(2.50, 1.5))
# Draw species-level effects from community means.
beta <- matrix(NA, nrow = N, ncol = p.occ)
alpha <- matrix(NA, nrow = N, ncol = p.det)
for (i in 1:p.occ) {
beta[, i] <- rnorm(N, beta.mean[i], sqrt(tau.sq.beta[i]))
}
for (i in 1:p.det) {
alpha[, i] <- rnorm(N, alpha.mean[i], sqrt(tau.sq.alpha[i]))
}
alpha.true <- alpha
phi <- rep(3 / .7, N)
sigma.sq <- rep(2, N)
nu <- rep(2, N)
dat <- simMsOcc(J.x = J.x, J.y = J.y, n.rep = n.rep, N = N, beta = beta, alpha = alpha,
psi.RE = psi.RE, p.RE = p.RE, sp = TRUE, sigma.sq = sigma.sq,
phi = phi, nu = nu, cov.model = 'matern')
pred.indx <- sample(1:J, round(J * .25), replace = FALSE)
y <- dat$y[, -pred.indx, , drop = FALSE]
# Occupancy covariates
X <- dat$X[-pred.indx, , drop = FALSE]
X.re <- dat$X.re[-pred.indx, , drop = FALSE]
coords <- as.matrix(dat$coords[-pred.indx, , drop = FALSE])
# Prediction covariates
X.0 <- dat$X[pred.indx, , drop = FALSE]
X.re.0 <- dat$X.re[pred.indx, , drop = FALSE]
coords.0 <- as.matrix(dat$coords[pred.indx, , drop = FALSE])
# Detection covariates
X.p <- dat$X.p[-pred.indx, , , drop = FALSE]
X.p.re <- dat$X.p.re[-pred.indx, , , drop = FALSE]
occ.covs <- cbind(X, X.re)
colnames(occ.covs) <- c('int', 'occ.cov.1', 'occ.cov.2', 'occ.factor.1')
det.covs <- list(det.cov.1 = X.p[, , 2],
det.factor.1 = X.p.re[, , 1],
det.factor.2 = X.p.re[, , 2])
data.list <- list(y = y, coords = coords, occ.covs = occ.covs, det.covs = det.covs)
# Priors
prior.list <- list(beta.comm.normal = list(mean = 0, var = 2.72),
alpha.comm.normal = list(mean = 0, var = 2.72),
tau.sq.beta.ig = list(a = 0.1, b = 0.1),
tau.sq.alpha.ig = list(a = 0.1, b = 0.1),
nu.unif = list(0.5, 2.5))
# Starting values
inits.list <- list(alpha.comm = 0,
beta.comm = 0,
beta = 0,
alpha = 0,
tau.sq.beta = 1,
tau.sq.alpha = 1,
z = apply(y, c(1, 2), max, na.rm = TRUE))
# Tuning
tuning.list <- list(phi = 1, nu = 0.25)
batch.length <- 25
n.batch <- 10
n.report <- 100
occ.formula <- ~ occ.cov.1 + occ.cov.2 + (1 | occ.factor.1)
det.formula <- ~ det.cov.1 + (1 | det.factor.1) + (1 | det.factor.2)
out <- spMsPGOcc(occ.formula = occ.formula,
det.formula = det.formula,
data = data.list,
inits = inits.list,
n.batch = n.batch,
batch.length = batch.length,
accept.rate = 0.43,
priors = prior.list,
cov.model = "matern",
tuning = tuning.list,
n.omp.threads = 1,
verbose = FALSE,
NNGP = TRUE,
n.neighbors = 5,
search.type = 'cb',
n.report = 10,
n.burn = 100,
n.thin = 2,
n.chains = 2,
k.fold = 2,
k.fold.threads = 1)
# To make sure it worked --------------
test_that("out is of class spMsPGOcc", {
expect_s3_class(out, "spMsPGOcc")
})
# Check cross-validation --------------
test_that("cross-validation works", {
expect_equal(length(out$k.fold.deviance), N)
expect_type(out$k.fold.deviance, "double")
expect_equal(sum(out$k.fold.deviance < 0), 0)
})
# Check random effects ----------------
test_that("random effects are correct", {
expect_equal(out$pRE, TRUE)
expect_equal(out$psiRE, TRUE)
})
# Check RE error ----------------------
test_that("random effect gives error when non-numeric", {
data.list$occ.covs <- as.data.frame(data.list$occ.covs)
data.list$occ.covs$occ.factor.1 <- factor(data.list$occ.covs$occ.factor.1)
data.list$det.covs$det.factor.1 <- factor(data.list$det.covs$det.factor.1)
expect_error(out <- spMsPGOcc(occ.formula = occ.formula,
det.formula = det.formula,
data = data.list,
inits = inits.list,
n.batch = n.batch,
batch.length = batch.length,
accept.rate = 0.43,
priors = prior.list,
cov.model = "matern",
tuning = tuning.list,
n.omp.threads = 1,
verbose = FALSE,
NNGP = TRUE,
n.neighbors = 5,
search.type = 'cb',
n.report = 10,
n.burn = 100,
n.thin = 2,
n.chains = 1))
data.list$occ.covs$occ.factor.1 <- as.character(factor(data.list$occ.covs$occ.factor.1))
data.list$det.covs$det.factor.1 <- as.character(factor(data.list$det.covs$det.factor.1))
expect_error(out <- spMsPGOcc(occ.formula = occ.formula,
det.formula = det.formula,
data = data.list,
inits = inits.list,
n.batch = n.batch,
batch.length = batch.length,
accept.rate = 0.43,
priors = prior.list,
cov.model = "matern",
tuning = tuning.list,
n.omp.threads = 1,
verbose = FALSE,
NNGP = TRUE,
n.neighbors = 5,
search.type = 'cb',
n.report = 10,
n.burn = 100,
n.thin = 2,
n.chains = 1))
})
# Check RE levels ---------------------
test_that("random effect levels are correct", {
expect_equal(sort(unique(unlist(out$re.level.names))),
sort(unique(c(X.re))))
expect_equal(sort(unique(unlist(out$p.re.level.names))),
sort(unique(c(X.p.re))))
})
# Check output data output is correct -
test_that("out$y == y", {
expect_equal(out$y, y)
})
# Check missing values ----------------
test_that("missing value error handling works", {
tmp.data <- data.list
tmp.data$occ.covs[3, ] <- NA
expect_error(spMsPGOcc(occ.formula = occ.formula,
det.formula = det.formula,
data = tmp.data,
n.batch = 40,
batch.length = batch.length,
cov.model = "exponential",
tuning = tuning.list,
NNGP = TRUE,
verbose = FALSE,
n.neighbors = 5,
search.type = 'cb',
n.report = 10,
n.burn = 100,
n.chains = 1))
tmp.data <- data.list
tmp.data$y[1, 1, 1] <- 1
tmp.data$det.covs[[1]][1] <- NA
expect_error(spMsPGOcc(occ.formula = occ.formula,
det.formula = det.formula,
data = tmp.data,
n.batch = 40,
batch.length = batch.length,
cov.model = "exponential",
tuning = tuning.list,
NNGP = TRUE,
verbose = FALSE,
n.neighbors = 5,
search.type = 'cb',
n.report = 10,
n.burn = 100,
n.chains = 1))
tmp.data <- data.list
tmp.data$y[, 1, 1] <- NA
out <- spMsPGOcc(occ.formula = occ.formula,
det.formula = det.formula,
data = tmp.data,
n.batch = 40,
batch.length = batch.length,
cov.model = "exponential",
tuning = tuning.list,
NNGP = TRUE,
verbose = FALSE,
n.neighbors = 5,
search.type = 'cb',
n.report = 10,
n.burn = 100,
n.chains = 1)
expect_s3_class(out, "spMsPGOcc")
})
# Check default values ----------------
test_that("default priors, inits, burn, thin work", {
out <- spMsPGOcc(occ.formula = occ.formula,
det.formula = det.formula,
data = data.list,
n.batch = n.batch,
batch.length = batch.length,
accept.rate = 0.43,
cov.model = "exponential",
tuning = tuning.list,
NNGP = TRUE,
verbose = FALSE,
n.neighbors = 5,
n.chains = 1)
expect_s3_class(out, "spMsPGOcc")
})
test_that("all correlation functions work", {
out <- spMsPGOcc(occ.formula = occ.formula,
det.formula = det.formula,
data = data.list,
inits = inits.list,
n.batch = n.batch,
batch.length = batch.length,
accept.rate = 0.43,
priors = prior.list,
cov.model = "gaussian",
tuning = list(phi = 0.3),
n.omp.threads = 1,
verbose = FALSE,
NNGP = TRUE,
n.neighbors = 5,
search.type = 'cb',
n.report = 10,
n.burn = 100,
n.thin = 1,
n.chains = 1)
expect_s3_class(out, "spMsPGOcc")
out <- spMsPGOcc(occ.formula = occ.formula,
det.formula = det.formula,
data = data.list,
inits = inits.list,
n.batch = n.batch,
batch.length = batch.length,
accept.rate = 0.43,
priors = prior.list,
cov.model = "spherical",
tuning = list(phi = 0.3),
n.omp.threads = 1,
verbose = FALSE,
NNGP = TRUE,
n.neighbors = 5,
search.type = 'cb',
n.report = 10,
n.burn = 100,
n.thin = 1,
n.chains = 1)
expect_s3_class(out, "spMsPGOcc")
out <- spMsPGOcc(occ.formula = occ.formula,
det.formula = det.formula,
data = data.list,
inits = inits.list,
n.batch = n.batch,
batch.length = batch.length,
accept.rate = 0.43,
priors = list(nu.unif = list(0.4, 3)),
cov.model = "matern",
tuning = list(phi = 0.3, nu = 0.2),
n.omp.threads = 1,
verbose = FALSE,
NNGP = TRUE,
n.neighbors = 5,
search.type = 'cb',
n.report = 10,
n.burn = 100,
n.thin = 1,
n.chains = 1)
expect_s3_class(out, "spMsPGOcc")
})
test_that("verbose prints to the screen", {
expect_output(spMsPGOcc(occ.formula = occ.formula,
det.formula = det.formula,
data = data.list,
inits = inits.list,
n.batch = n.batch,
batch.length = batch.length,
accept.rate = 0.43,
priors = prior.list,
cov.model = "exponential",
tuning = tuning.list,
n.omp.threads = 1,
verbose = TRUE,
NNGP = TRUE,
n.neighbors = 5,
search.type = 'cb',
n.report = 10,
n.burn = 100,
n.thin = 1,
n.chains = 1))
})
# Check waicOcc -----------------------
test_that("waicOCC works for spMsPGOcc", {
# as.vector gets rid of names
waic.out <- as.vector(waicOcc(out))
expect_equal(length(waic.out), 3)
expect_equal(waic.out[3], -2 * (waic.out[1] - waic.out[2]))
})
test_that("fitted works for spMsPGOcc", {
fitted.out <- fitted(out)
expect_equal(length(fitted.out), 2)
expect_equal(class(fitted.out$y.rep.samples), "array")
expect_equal(class(fitted.out$p.samples), "array")
expect_equal(dim(fitted.out$y.rep.samples), dim(fitted.out$p.samples))
})
test_that("predict works for spMsPGOcc", {
n.post.samples <- out$n.post * out$n.chains
X.0 <- cbind(X.0, X.re.0)
colnames(X.0) <- c('int', 'occ.cov.1', 'occ.cov.2', 'occ.factor.1')
pred.out <- predict(out, X.0, coords.0, verbose = FALSE)
expect_type(pred.out, "list")
expect_equal(dim(pred.out$psi.0.samples), c(n.post.samples, N, nrow(X.0)))
expect_equal(dim(pred.out$z.0.samples), c(n.post.samples, N, nrow(X.0)))
})
test_that("detection prediction works", {
X.p.0 <- cbind(dat$X.p[, 1, ], dat$X.p.re[, 1, 1:2])
colnames(X.p.0) <- c('intercept', 'det.cov.1', 'det.factor.1', 'det.factor.2')
pred.out <- predict(out, X.p.0, type = 'detection')
expect_type(pred.out, 'list')
expect_equal(dim(pred.out$p.0.samples), c(out$n.post * out$n.chains, N, J))
})
test_that("posterior predictive checks work for msPGOcc", {
n.post.samples <- out$n.post * out$n.chains
J.fit <- nrow(X)
ppc.out <- ppcOcc(out, 'chi-square', 2)
expect_type(ppc.out, "list")
expect_equal(dim(ppc.out$fit.y), c(n.post.samples, N))
expect_equal(dim(ppc.out$fit.y.rep), c(n.post.samples, N))
expect_equal(dim(ppc.out$fit.y.group.quants), c(5, N, max(n.rep)))
expect_equal(dim(ppc.out$fit.y.rep.group.quants), c(5, N, max(n.rep)))
ppc.out <- ppcOcc(out, 'chi-square', 1)
expect_type(ppc.out, "list")
expect_equal(dim(ppc.out$fit.y), c(n.post.samples, N))
expect_equal(dim(ppc.out$fit.y.rep), c(n.post.samples, N))
expect_equal(dim(ppc.out$fit.y.group.quants), c(5, N, J.fit))
expect_equal(dim(ppc.out$fit.y.rep.group.quants), c(5, N, J.fit))
ppc.out <- ppcOcc(out, 'freeman-tukey', 1)
expect_type(ppc.out, "list")
expect_equal(dim(ppc.out$fit.y), c(n.post.samples, N))
expect_equal(dim(ppc.out$fit.y.rep), c(n.post.samples, N))
expect_equal(dim(ppc.out$fit.y.group.quants), c(5, N, J.fit))
expect_equal(dim(ppc.out$fit.y.rep.group.quants), c(5, N, J.fit))
ppc.out <- ppcOcc(out, 'freeman-tukey', 2)
expect_type(ppc.out, "list")
expect_equal(dim(ppc.out$fit.y), c(n.post.samples, N))
expect_equal(dim(ppc.out$fit.y.rep), c(n.post.samples, N))
expect_equal(dim(ppc.out$fit.y.group.quants), c(5, N, max(n.rep)))
expect_equal(dim(ppc.out$fit.y.rep.group.quants), c(5, N, max(n.rep)))
})
# Model with fixed sigma.sq -----------------------------------------------
test_that("spMsPGOcc works with fixed sigma.sq", {
out <- spMsPGOcc(occ.formula = occ.formula,
det.formula = det.formula,
data = data.list,
inits = inits.list,
n.batch = n.batch,
batch.length = batch.length,
accept.rate = 0.43,
priors = list(sigma.sq.ig = "fixed"),
cov.model = "exponential",
tuning = list(phi = 0.3, nu = 0.2),
n.omp.threads = 1,
verbose = FALSE,
NNGP = TRUE,
n.neighbors = 5,
search.type = 'cb',
n.report = 10,
n.burn = 100,
n.thin = 1,
n.chains = 1)
expect_s3_class(out, "spMsPGOcc")
expect_equal(length(unique(out$theta.samples[, 1])), 1)
})
# Uniform sigma sq --------------------------------------------------------
test_that("spMsPGOcc works with uniform prior on sigma.sq", {
prior.list <- list(sigma.sq.unif = list(a = 0, b = 5),
nu.unif = list(a = 0.1, b = 4))
tuning.list <- list(phi = 0.5, nu = 0.6, sigma.sq = 0.7)
out <- spMsPGOcc(occ.formula = occ.formula,
det.formula = det.formula,
data = data.list,
n.batch = 40,
batch.length = batch.length,
cov.model = "exponential",
priors = prior.list,
tuning = tuning.list,
NNGP = TRUE,
verbose = FALSE,
n.neighbors = 5,
search.type = 'cb',
n.report = 10,
n.burn = 100,
n.chains = 1)
expect_s3_class(out, "spMsPGOcc")
out <- spMsPGOcc(occ.formula = occ.formula,
det.formula = det.formula,
data = data.list,
n.batch = 40,
batch.length = batch.length,
cov.model = "matern",
priors = prior.list,
tuning = tuning.list,
NNGP = TRUE,
verbose = FALSE,
n.neighbors = 5,
search.type = 'cb',
n.report = 10,
n.burn = 100,
n.chains = 1)
expect_s3_class(out, "spMsPGOcc")
})
# Third dimension of y != max(n.rep)
J.x <- 8
J.y <- 8
J <- J.x * J.y
n.rep<- sample(2:4, size = J, replace = TRUE)
n.rep.max <- 7
N <- 6
# Community-level covariate effects
# Occurrence
beta.mean <- c(0.2, 0.5, 1.2)
p.occ <- length(beta.mean)
tau.sq.beta <- c(0.6, 1.5, 2.3)
# Detection
alpha.mean <- c(0, -0.5)
tau.sq.alpha <- c(1, 2.3)
p.det <- length(alpha.mean)
# Random effects
psi.RE <- list()
p.RE <- list()
# Draw species-level effects from community means.
beta <- matrix(NA, nrow = N, ncol = p.occ)
alpha <- matrix(NA, nrow = N, ncol = p.det)
for (i in 1:p.occ) {
beta[, i] <- rnorm(N, beta.mean[i], sqrt(tau.sq.beta[i]))
}
for (i in 1:p.det) {
alpha[, i] <- rnorm(N, alpha.mean[i], sqrt(tau.sq.alpha[i]))
}
alpha.true <- alpha
phi <- rep(3 / .7, N)
sigma.sq <- rep(2, N)
nu <- rep(2, N)
dat <- simMsOcc(J.x = J.x, J.y = J.y, n.rep = n.rep, N = N, beta = beta, alpha = alpha,
psi.RE = psi.RE, p.RE = p.RE, sp = TRUE, sigma.sq = sigma.sq,
phi = phi, nu = nu, cov.model = 'matern', n.rep.max = n.rep.max)
pred.indx <- sample(1:J, round(J * .25), replace = FALSE)
y <- dat$y[, -pred.indx, , drop = FALSE]
# Occupancy covariates
X <- dat$X[-pred.indx, , drop = FALSE]
coords <- as.matrix(dat$coords[-pred.indx, , drop = FALSE])
# Prediction covariates
X.0 <- dat$X[pred.indx, , drop = FALSE]
coords.0 <- as.matrix(dat$coords[pred.indx, , drop = FALSE])
# Detection covariates
X.p <- dat$X.p[-pred.indx, , , drop = FALSE]
occ.covs <- X
colnames(occ.covs) <- c('int', 'occ.cov.1', 'occ.cov.2')
det.covs <- list(det.cov.1 = X.p[, , 2])
data.list <- list(y = y, coords = coords, occ.covs = occ.covs, det.covs = det.covs)
# Priors
prior.list <- list(beta.comm.normal = list(mean = 0, var = 2.72),
alpha.comm.normal = list(mean = 0, var = 2.72),
tau.sq.beta.ig = list(a = 0.1, b = 0.1),
tau.sq.alpha.ig = list(a = 0.1, b = 0.1),
nu.unif = list(0.5, 2.5))
# Starting values
inits.list <- list(alpha.comm = 0,
beta.comm = 0,
beta = 0,
alpha = 0,
tau.sq.beta = 1,
tau.sq.alpha = 1,
z = apply(y, c(1, 2), max, na.rm = TRUE))
# Tuning
tuning.list <- list(phi = 1, nu = 0.25)
batch.length <- 25
n.batch <- 10
n.report <- 100
occ.formula <- ~ occ.cov.1 + occ.cov.2
det.formula <- ~ det.cov.1
out <- spMsPGOcc(occ.formula = occ.formula,
det.formula = det.formula,
data = data.list,
inits = inits.list,
n.batch = n.batch,
batch.length = batch.length,
accept.rate = 0.43,
priors = prior.list,
cov.model = "matern",
tuning = tuning.list,
n.omp.threads = 1,
verbose = FALSE,
NNGP = TRUE,
n.neighbors = 5,
search.type = 'cb',
n.report = 10,
n.burn = 100,
n.thin = 2,
n.chains = 2,
k.fold = 2,
k.fold.threads = 1)
# To make sure it worked --------------
test_that("out is of class spMsPGOcc", {
expect_s3_class(out, "spMsPGOcc")
})
# Check cross-validation --------------
test_that("cross-validation works", {
expect_equal(length(out$k.fold.deviance), N)
expect_type(out$k.fold.deviance, "double")
expect_equal(sum(out$k.fold.deviance < 0), 0)
})
# Check random effects ----------------
test_that("random effects are correct", {
expect_equal(out$pRE, FALSE)
expect_equal(out$psiRE, FALSE)
})
# Check output data output is correct -
test_that("out$y == y", {
expect_equal(out$y, y)
})
# Check missing values ----------------
test_that("missing value error handling works", {
tmp.data <- data.list
tmp.data$occ.covs[3, ] <- NA
expect_error(spMsPGOcc(occ.formula = occ.formula,
det.formula = det.formula,
data = tmp.data,
n.batch = 40,
batch.length = batch.length,
cov.model = "exponential",
tuning = tuning.list,
NNGP = TRUE,
verbose = FALSE,
n.neighbors = 5,
search.type = 'cb',
n.report = 10,
n.burn = 100,
n.chains = 1))
tmp.data <- data.list
tmp.data$y[1, 1, 1] <- 1
tmp.data$det.covs[[1]][1] <- NA
expect_error(spMsPGOcc(occ.formula = occ.formula,
det.formula = det.formula,
data = tmp.data,
n.batch = 40,
batch.length = batch.length,
cov.model = "exponential",
tuning = tuning.list,
NNGP = TRUE,
verbose = FALSE,
n.neighbors = 5,
search.type = 'cb',
n.report = 10,
n.burn = 100,
n.chains = 1))
tmp.data <- data.list
tmp.data$y[, 1, 1] <- NA
out <- spMsPGOcc(occ.formula = occ.formula,
det.formula = det.formula,
data = tmp.data,
n.batch = 40,
batch.length = batch.length,
cov.model = "exponential",
tuning = tuning.list,
NNGP = TRUE,
verbose = FALSE,
n.neighbors = 5,
search.type = 'cb',
n.report = 10,
n.burn = 100,
n.chains = 1)
expect_s3_class(out, "spMsPGOcc")
})
# Check default values ----------------
test_that("default priors, inits, burn, thin work", {
out <- spMsPGOcc(occ.formula = occ.formula,
det.formula = det.formula,
data = data.list,
n.batch = n.batch,
batch.length = batch.length,
accept.rate = 0.43,
cov.model = "exponential",
tuning = tuning.list,
NNGP = TRUE,
verbose = FALSE,
n.neighbors = 5,
n.chains = 1)
expect_s3_class(out, "spMsPGOcc")
})
test_that("all correlation functions work", {
out <- spMsPGOcc(occ.formula = occ.formula,
det.formula = det.formula,
data = data.list,
inits = inits.list,
n.batch = n.batch,
batch.length = batch.length,
accept.rate = 0.43,
priors = prior.list,
cov.model = "gaussian",
tuning = list(phi = 0.3),
n.omp.threads = 1,
verbose = FALSE,
NNGP = TRUE,
n.neighbors = 5,
search.type = 'cb',
n.report = 10,
n.burn = 100,
n.thin = 1,
n.chains = 1)
expect_s3_class(out, "spMsPGOcc")
out <- spMsPGOcc(occ.formula = occ.formula,
det.formula = det.formula,
data = data.list,
inits = inits.list,
n.batch = n.batch,
batch.length = batch.length,
accept.rate = 0.43,
priors = prior.list,
cov.model = "spherical",
tuning = list(phi = 0.3),
n.omp.threads = 1,
verbose = FALSE,
NNGP = TRUE,
n.neighbors = 5,
search.type = 'cb',
n.report = 10,
n.burn = 100,
n.thin = 1,
n.chains = 1)
expect_s3_class(out, "spMsPGOcc")
out <- spMsPGOcc(occ.formula = occ.formula,
det.formula = det.formula,
data = data.list,
inits = inits.list,
n.batch = n.batch,
batch.length = batch.length,
accept.rate = 0.43,
priors = list(nu.unif = list(0.4, 3)),
cov.model = "matern",
tuning = list(phi = 0.3, nu = 0.2),
n.omp.threads = 1,
verbose = FALSE,
NNGP = TRUE,
n.neighbors = 5,
search.type = 'cb',
n.report = 10,
n.burn = 100,
n.thin = 1,
n.chains = 1)
expect_s3_class(out, "spMsPGOcc")
})
test_that("verbose prints to the screen", {
expect_output(spMsPGOcc(occ.formula = occ.formula,
det.formula = det.formula,
data = data.list,
inits = inits.list,
n.batch = n.batch,
batch.length = batch.length,
accept.rate = 0.43,
priors = prior.list,
cov.model = "exponential",
tuning = tuning.list,
n.omp.threads = 1,
verbose = TRUE,
NNGP = TRUE,
n.neighbors = 5,
search.type = 'cb',
n.report = 10,
n.burn = 100,
n.thin = 1,
n.chains = 1))
})
# Check waicOcc -----------------------
test_that("waicOCC works for spMsPGOcc", {
# as.vector gets rid of names
waic.out <- as.vector(waicOcc(out))
expect_equal(length(waic.out), 3)
expect_equal(waic.out[3], -2 * (waic.out[1] - waic.out[2]))
})
test_that("fitted works for spMsPGOcc", {
fitted.out <- fitted(out)
expect_equal(length(fitted.out), 2)
expect_equal(class(fitted.out$y.rep.samples), "array")
expect_equal(class(fitted.out$p.samples), "array")
expect_equal(dim(fitted.out$y.rep.samples), dim(fitted.out$p.samples))
})
test_that("predict works for spMsPGOcc", {
n.post.samples <- out$n.post * out$n.chains
pred.out <- predict(out, X.0, coords.0, verbose = FALSE)
expect_type(pred.out, "list")
expect_equal(dim(pred.out$psi.0.samples), c(n.post.samples, N, nrow(X.0)))
expect_equal(dim(pred.out$z.0.samples), c(n.post.samples, N, nrow(X.0)))
})
test_that("detection prediction works", {
X.p.0 <- dat$X.p[, 1, ]
pred.out <- predict(out, X.p.0, type = 'detection')
expect_type(pred.out, 'list')
expect_equal(dim(pred.out$p.0.samples), c(out$n.post * out$n.chains, N, J))
})
test_that("posterior predictive checks work for msPGOcc", {
n.post.samples <- out$n.post * out$n.chains
J.fit <- nrow(X)
ppc.out <- ppcOcc(out, 'chi-square', 2)
expect_type(ppc.out, "list")
expect_equal(dim(ppc.out$fit.y), c(n.post.samples, N))
expect_equal(dim(ppc.out$fit.y.rep), c(n.post.samples, N))
expect_equal(dim(ppc.out$fit.y.group.quants), c(5, N, n.rep.max))
expect_equal(dim(ppc.out$fit.y.rep.group.quants), c(5, N, n.rep.max))
ppc.out <- ppcOcc(out, 'chi-square', 1)
expect_type(ppc.out, "list")
expect_equal(dim(ppc.out$fit.y), c(n.post.samples, N))
expect_equal(dim(ppc.out$fit.y.rep), c(n.post.samples, N))
expect_equal(dim(ppc.out$fit.y.group.quants), c(5, N, J.fit))
expect_equal(dim(ppc.out$fit.y.rep.group.quants), c(5, N, J.fit))
ppc.out <- ppcOcc(out, 'freeman-tukey', 1)
expect_type(ppc.out, "list")
expect_equal(dim(ppc.out$fit.y), c(n.post.samples, N))
expect_equal(dim(ppc.out$fit.y.rep), c(n.post.samples, N))
expect_equal(dim(ppc.out$fit.y.group.quants), c(5, N, J.fit))
expect_equal(dim(ppc.out$fit.y.rep.group.quants), c(5, N, J.fit))
ppc.out <- ppcOcc(out, 'freeman-tukey', 2)
expect_type(ppc.out, "list")
expect_equal(dim(ppc.out$fit.y), c(n.post.samples, N))
expect_equal(dim(ppc.out$fit.y.rep), c(n.post.samples, N))
expect_equal(dim(ppc.out$fit.y.group.quants), c(5, N, n.rep.max))
expect_equal(dim(ppc.out$fit.y.rep.group.quants), c(5, N, n.rep.max))
})
R Package Documentation
Browse R Packages
We want your feedback!
Note that we can't provide technical support on individual packages. You should contact the package authors for that.
Embedding an R snippet on your website
Add the following code to your website.
For more information on customizing the embed code, read Embedding Snippets.