tests/testthat/test_FLXMCL_svm.R
In schiffner/locClass: Collection of Local Classification Methods
context("FLXMCLsvm")
test_that("FLXMCLsvm: misspecified arguments", {
# wrong variable names
cluster <- kmeans(iris[,1:4], centers = 3)$cluster
expect_that(fit <- flexmix(Species ~ V1, data = iris, concomitant = FLXPmultinom(~ Sepal.Length), model = FLXMCLsvm(kernel = "linear", fitted = FALSE), cluster = cluster, control = list(iter.max = 200, classify = "hard")), throws_error("object 'V1' not found"))
expect_that(fit <- flexmix(Species ~ Sepal.Length, data = iris, concomitant = FLXPmultinom(~ V1), model = FLXMCLsvm(kernel = "linear", fitted = FALSE), cluster = cluster, control = list(iter.max = 200, classify = "hard")), throws_error("object 'V1' not found"))
expect_that(fit <- flexmix(y ~ Sepal.Length, data = iris, concomitant = FLXPmultinom(~ Sepal.Width), model = FLXMCLsvm(kernel = "linear", fitted = FALSE), cluster = cluster, control = list(iter.max = 200, classify = "hard")), throws_error("object 'y' not found"))
# wrong class
expect_error(fit <- flexmix(iris, data = iris, concomitant = FLXPmultinom(~ Sepal.Width), model = FLXMCLsvm(kernel = "linear", fitted = FALSE), cluster = cluster, control = list(iter.max = 200, classify = "hard")))
# target variable also in x
# expect_error(mob(Species ~ Species + Sepal.Length | Sepal.Width, data = iris, model = majorityModel,
# control = mob_control(objfun = deviance, minsplit = 20))) ## funktioniert, sollte aber nicht
})
test_that("FLXMCLsvm: arguments are passed to wsvm",{
iris[,1:4] <- scale(iris[,1:4])
set.seed(123)
cluster <- kmeans(iris[,"Sepal.Width"], centers = 3)$cluster
## kernel, cost
fit <- flexmix(Species ~ Sepal.Width, data = iris, model = FLXMCLsvm(kernel = "linear", cost = 2), cluster = cluster, control = list(iter.max = 50, classify = "weighted", verb = 1))
expect_equal(fit@components[[1]][[1]]@parameters$kernel, 0)
expect_equal(fit@components[[1]][[1]]@parameters$cost, 2)
expect_true(fit@components[[1]][[1]]@parameters$fitted)
## degree, coef0
fit <- flexmix(Species ~ Sepal.Width, data = iris, model = FLXMCLsvm(kernel = "polynomial", coef0 = 2, degree = 2, fitted = TRUE, tolerance = 10^-5), cluster = cluster, control = list(iter.max = 50, classify = "weighted", verb = 1))
expect_equal(fit@components[[1]][[1]]@parameters$kernel, 1)
expect_equal(fit@components[[1]][[1]]@parameters$coef0, 2)
expect_equal(fit@components[[1]][[1]]@parameters$degree, 2)
expect_true(fit@components[[1]][[1]]@parameters$fitted)
## gamma
fit <- flexmix(Species ~ Sepal.Width, data = iris, model = FLXMCLsvm(kernel = "radial", gamma = 2, fitted = FALSE, tolerance = 10^-5), cluster = cluster, control = list(iter.max = 50, classify = "weighted", verb = 1))
expect_equal(fit@components[[1]][[1]]@parameters$kernel, 2)
expect_equal(fit@components[[1]][[1]]@parameters$gamma, 2)
expect_true(!fit@components[[1]][[1]]@parameters$fitted)
})
test_that("FLXMCLsvm with several options works",{
cluster <- kmeans(iris[,1], centers = 2)$cluster
# ## weighted, FLXPwlda
# fit <- flexmix(Species ~ Sepal.Width, data = iris, concomitant = FLXPwlda(~ Sepal.Length), model = FLXMCLsvm(kernel = "linear", fitted = FALSE), cluster = cluster, control = list(iter.max = 200, classify = "weighted", verb = 1))
# # ok
# ## hard, FLXPwlda
# fit <- flexmix(Species ~ Sepal.Width, data = iris, concomitant = FLXPwlda(~ Sepal.Length), model = FLXMCLsvm(kernel = "linear", fitted = FALSE), cluster = cluster, control = list(iter.max = 200, classify = "hard", verb = 1))
# # not monotone, empty classes
## weighted, FLXPmultinom
fit <- flexmix(Species ~ Sepal.Width, data = iris, concomitant = FLXPmultinom(~ Sepal.Length), model = FLXMCLsvm(kernel = "linear", fitted = FALSE), cluster = cluster, control = list(iter.max = 200, classify = "weighted", verb = 1))
# ok
## hard, FLXPmultinom
fit <- flexmix(Species ~ Sepal.Width, data = iris, concomitant = FLXPmultinom(~ Sepal.Length), model = FLXMCLsvm(kernel = "linear", fitted = FALSE), cluster = cluster, control = list(iter.max = 200, classify = "hard", verb = 1))
# ok
})
test_that("FLXMCLsvm throws a warning if grouping variable is numeric", {
iris$Species <- as.numeric(iris$Species)
cluster <- kmeans(iris[,c(1,3)], centers = 2)$cluster
expect_that(tr2 <- flexmix(Species ~ Sepal.Length, data = iris, concomitant = FLXPmultinom(~ Petal.Length + Sepal.Length), model = FLXMCLsvm(kernel = "linear", fitted = FALSE, tolerance = 10^-6), cluster = cluster, control = list(iter.max = 50, classify = "weighted", tolerance = 10^-4, verb = 1)), gives_warning("'grouping' was coerced to a factor"))
## use weighted here instead of hard, otherwise SVM does not work
})
test_that("FLXMCLsvm works if only one predictor variable is given", {
cluster <- kmeans(iris[,1:4], centers = 2)$cluster
fit <- flexmix(Species ~ Sepal.Width, data = iris, concomitant = FLXPmultinom(~ Sepal.Length), model = FLXMCLsvm(kernel = "linear", fitted = FALSE), cluster = cluster, control = list(iter.max = 200, classify = "weighted", verb = 1))
})
test_that("FLXMCLsvm: Local and global solution coincide if only one cluster is given", {
fit <- flexmix(Species ~ ., data = iris, model = FLXMCLsvm(kernel = "linear", fitted = FALSE), cluster = 1, control = list(iter.max = 200, classify = "hard"))
w <- wsvm(Species ~ ., data = iris, kernel = "linear")
expect_equal(fit@components[[1]][[1]]@parameters$coefs, w$coefs)
pred <- max.col(mypredict(fit)[[1]])
p <- predict(w)
expect_equal(pred, as.numeric(p))
})
test_that("FLXMCLsvm: training data from only one class", {
cluster <- kmeans(iris[1:50,1:4], centers = 3)$cluster
expect_that(fit <- flexmix(Species ~ Sepal.Width + Sepal.Length, data = iris[1:50,], concomitant = FLXPmultinom(~ Sepal.Width + Sepal.Length), model = FLXMCLsvm(kernel = "linear", fitted = FALSE), cluster = cluster, control = list(iter.max = 200, classify = "hard")), throws_error("need training data from at least two classes"))
})
test_that("FLXMCLsvm: missing classes in individual clusters", {
set.seed(123)
cluster <- kmeans(iris[,1], centers = 2)$cluster
expect_that(fit <- flexmix(Species ~ Sepal.Width, data = iris, concomitant = FLXPmultinom(~ Sepal.Length), model = FLXMCLsvm(kernel = "linear", fitted = FALSE), cluster = cluster, control = list(iter.max = 100, classify = "hard")), gives_warning("some groups are empty"))
expect_equal(ncol(fit@components$Comp.1[[1]]@parameters$coefs), 2)
expect_equal(ncol(fit@components$Comp.2[[1]]@parameters$coefs), 1)
pred <- mypredict(fit, aggregate = FALSE)
expect_equal(colnames(pred$Comp.1), levels(iris$Species))
expect_equal(colnames(pred$Comp.2), levels(iris$Species))
expect_true(all(pred$Comp.2[,1] == 0))
irisna <- iris
irisna$Sepal.Width[1:12] <- NA
pred <- mypredict(fit, newdata = irisna, aggregate = FALSE)
expect_equal(colnames(pred$Comp.1), levels(iris$Species))
expect_equal(colnames(pred$Comp.2), levels(iris$Species))
expect_true(all(pred$Comp.2[-c(1:12),1] == 0))
# library(mlbench)
# data(Glass)
# set.seed(120)
# cluster <- kmeans(Glass[,1], centers = 2)$cluster
# expect_that(fit <- flexmix(Type ~ Na, data = Glass, concomitant = FLXPwlda(as.formula(paste("~", paste(colnames(Glass)[1:9], collapse = "+")))), model = FLXMCLsvm(kernel = "linear", fitted = FALSE), cluster = cluster, control = list(iter.max = 200, classify = "hard", verb = 1)), gives_warning("some groups are empty"))
# expect_equal(ncol(fit@components$Comp.1[[1]]@parameters$coefs), 4)
# expect_equal(ncol(fit@components$Comp.2[[1]]@parameters$coefs), 5)
# pred <- mypredict(fit, aggregate = FALSE)
# expect_equal(colnames(pred$Comp.1), as.character(c(1:3,5:7)))
# expect_equal(colnames(pred$Comp.2), as.character(c(1:3,5:7)))
# expect_equal(colnames(pred$Comp.1[[2]]), c("1/2", "1/3", "1/5", "1/6", "1/7", "2/3", "2/5", "2/6", "2/7", "3/5", "3/6", "3/7", "5/6", "5/7", "6/7"))
# expect_equal(colnames(pred$Comp.2[[2]]), c("1/2", "1/3", "1/5", "1/6", "1/7", "2/3", "2/5", "2/6", "2/7", "3/5", "3/6", "3/7", "5/6", "5/7", "6/7"))
# expect_true(all(pred$Comp.2[[1]][,5] == 0))
# expect_true(all(pred$Comp.2[[2]][,c(4,8,11,13,15)] == 0))
# Glassna <- Glass
# Glassna$Na[1:12] <- NA
# pred <- mypredict(fit, newdata = Glassna, aggregate = FALSE)
# expect_equal(colnames(pred$Comp.1[[1]]), as.character(c(1:3,5:7)))
# expect_equal(colnames(pred$Comp.2[[1]]), as.character(c(1:3,5:7)))
# expect_equal(colnames(pred$Comp.1[[2]]), c("1/2", "1/3", "1/5", "1/6", "1/7", "2/3", "2/5", "2/6", "2/7", "3/5", "3/6", "3/7", "5/6", "5/7", "6/7"))
# expect_equal(colnames(pred$Comp.2[[2]]), c("1/2", "1/3", "1/5", "1/6", "1/7", "2/3", "2/5", "2/6", "2/7", "3/5", "3/6", "3/7", "5/6", "5/7", "6/7"))
# expect_true(all(pred$Comp.2[[1]][-c(1:12),5] == 0))
# expect_true(all(pred$Comp.2[[2]][-c(1:12),c(4,8,11,13,15)] == 0))
# expect_true(all(is.na(pred$Comp.1[[1]][1:12,])))
# expect_true(all(is.na(pred$Comp.1[[2]][1:12,])))
# expect_true(all(is.na(pred$Comp.2[[1]][1:12,])))
# expect_true(all(is.na(pred$Comp.2[[2]][1:12,])))
})
test_that("FLXMCLsvm: removing clusters works", {
set.seed(120)
data <- benchData::flashData(500)
cluster <- kmeans(data$x, centers = 12)$cluster
tr2 <- flexmix(y ~ ., data = as.data.frame(data), concomitant = FLXPmultinom(~ x.1 + x.2), model = FLXMCLsvm(tolerance = 10^-6), cluster = cluster, control = list(iter.max = 100, verb = 1))
expect_equal(length(tr2@components), 9)
expect_equal(ncol(tr2@posterior$scaled), 9)
})
#=================================================================================================================
context("predict FLXMCLsvm")
test_that("predict FLXMCLsvm works correctly with missing newdata", {
set.seed(120)
ran <- sample(1:500,300)
data <- benchData::flashData(500)
cluster <- kmeans(data$x[ran,], centers = 2)$cluster
tr2 <- flexmix(y ~ ., data = as.data.frame(data)[ran,], concomitant = FLXPmultinom(~ x.1 + x.2), model = FLXMCLsvm(kernel = "linear", fitted = FALSE), cluster = cluster, control = list(iter.max = 200))
pred1 <- mypredict(tr2, aggregate = FALSE)
pred2 <- mypredict(tr2, aggregate = FALSE, newdata = as.data.frame(data)[ran,])
expect_equal(pred1, pred2)
expect_equal(rownames(pred1$Comp.1), rownames(as.data.frame(data)[ran,]))
expect_equal(rownames(pred1$Comp.1), rownames(as.data.frame(data)[ran,]))
pred1 <- mypredict(tr2, aggregate = TRUE)
pred2 <- mypredict(tr2, aggregate = TRUE, newdata = as.data.frame(data)[ran,])
expect_equal(pred1, pred2)
expect_equal(rownames(pred1[[1]]), rownames(as.data.frame(data)[ran,]))
})
test_that("predict FLXMCLsvm works with missing classes in the training data", {
ran <- sample(1:150,100)
cluster <- kmeans(iris[1:100,1:4], centers = 2)$cluster
expect_that(tr2 <- flexmix(Species ~ ., data = iris[1:100,], concomitant = FLXPmultinom(~ Sepal.Width + Sepal.Length), model = FLXMCLsvm(), cluster = cluster, control = list(iter.max = 200)), gives_warning("some groups are empty"))
pred <- mypredict(tr2, aggregate = FALSE)
expect_equal(ncol(pred[[1]]), 3) #!!!
expect_equal(ncol(pred[[2]]), 3) #!!!
expect_true(all(pred[[1]][,3] == 0))
expect_true(all(pred[[2]][,3] == 0))
pred <- mypredict(tr2, aggregate = TRUE)
expect_equal(ncol(pred[[1]]), 3) #!!!
expect_true(all(pred[[1]][,3] == 0))
})
test_that("predict FLXMCLsvm works with one single predictor variable", {
ran <- sample(1:150,100)
cluster <- kmeans(iris[ran,1:4], centers = 2)$cluster
tr2 <- flexmix(Species ~ Sepal.Width, data = iris[ran,], concomitant = FLXPmultinom(~ Sepal.Width), model = FLXMCLsvm(kernel = "linear", fitted = FALSE), cluster = cluster, control = list(iter.max = 200))
pred <- mypredict(tr2, newdata = iris[-ran,], aggregate = FALSE)
pred <- mypredict(tr2, aggregate = TRUE)
})
test_that("predict FLXMCLsvm works with one single test observation", {
set.seed(120)
ran <- sample(1:150,100)
cluster <- kmeans(iris[ran,1:4], centers = 2)$cluster
tr2 <- flexmix(Species ~ Sepal.Width, data = iris[ran,], concomitant = FLXPmultinom(~ Sepal.Width), model = FLXMCLsvm(kernel = "linear", fitted = FALSE), cluster = cluster, control = list(iter.max = 200))
pred <- mypredict(tr2, newdata = iris[5,])
expect_equal(dim(pred[[1]]), c(1,3))
expect_equal(dim(pred[[2]]), c(1,3))
pred <- mypredict(tr2, newdata = iris[5,], aggregate = TRUE)
expect_equal(dim(pred[[1]]), c(1,3))
})
test_that("predict FLXMCLsvm: NA handling in newdata works", {
ran <- sample(1:150,100)
cluster <- kmeans(iris[ran,1:4], centers = 2)$cluster
tr2 <- flexmix(Species ~ Sepal.Length + Sepal.Width, data = iris[ran,], concomitant = FLXPmultinom(~ Sepal.Width + Petal.Width), model = FLXMCLsvm(kernel = "linear", fitted = FALSE), cluster = cluster, control = list(iter.max = 200))
## NAs in explanatory variables are ok
irisna <- iris
irisna[1:17,c(1,3)] <- NA
pred <- mypredict(tr2, newdata = irisna)
expect_equal(all(is.na(pred[[1]][1:17,])), TRUE)
expect_equal(all(is.na(pred[[2]][1:17,])), TRUE)
pred <- mypredict(tr2, newdata = irisna, aggregate = TRUE)
expect_equal(all(is.na(pred[[1]][1:17,])), TRUE)
## NAs in splitting variable are not ok if aggregation is desired
irisna[1:17,1:3] <- NA
pred <- mypredict(tr2, newdata = irisna)
expect_equal(all(is.na(pred[[1]][1:17,])), TRUE)
expect_equal(all(is.na(pred[[2]][1:17,])), TRUE)
pred <- mypredict(tr2, newdata = irisna, aggregate = TRUE)
expect_equal(all(is.na(pred[[1]][1:17,])), TRUE)
})
test_that("predict FLXMCLsvm: misspecified arguments", {
ran <- sample(1:150,100)
cluster <- kmeans(iris[ran,1:4], centers = 2)$cluster
tr2 <- flexmix(Species ~ Sepal.Width + Petal.Width, data = iris[ran,], concomitant = FLXPmultinom(~ Sepal.Width + Petal.Width), model = FLXMCLsvm(), cluster = cluster, control = list(iter.max = 200))
# errors in newdata
expect_error(mypredict(tr2, newdata = TRUE))
expect_error(mypredict(tr2, newdata = -50:50))
})
#=================================================================================================================
# library(locClass)
# d <- flashData(500)
# #d <- vNormalData(500)
# grid <- expand.grid(x.1=seq(-6,6,0.2), x.2=seq(-4,4,0.2))
# model <- FLXMCLsvm(kernel = "linear")
# cluster <- kmeans(d$x, center = 2)$cluster
# res <- flexmix(y ~ ., data = as.data.frame(d), concomitant = FLXPmultinom(~ x.1 + x.2), model = model, cluster = cluster)
# res
# plot(d$x, col = res@cluster, cex = res@posterior$scaled[,1])
# plot(d$x, col = res@cluster, cex = res@posterior$scaled[,2])
# plot(d$x, col = d$y, cex = res@posterior$scaled[,1])
# plot(d$x, col = d$y, cex = res@posterior$scaled[,2])
# pred <- predict(res, newdata = as.data.frame(d), local.aggregate = TRUE)
# pred.grid <- predict(res, newdata = grid)
# image(seq(-6,6,0.2), seq(-4,4,0.2), matrix(pred.grid$Comp.1["decision",][[1]], length(seq(-6,6,0.2))))
# contour(seq(-6,6,0.2), seq(-4,4,0.2), matrix(pred.grid$Comp.1["decision",][[1]], length(seq(-6,6,0.2))), add = TRUE)
# image(seq(-6,6,0.2), seq(-4,4,0.2), matrix(pred.grid$Comp.2["decision",][[1]], length(seq(-6,6,0.2))))
# contour(seq(-6,6,0.2), seq(-4,4,0.2), matrix(pred.grid$Comp.2["decision",][[1]], length(seq(-6,6,0.2))), add = TRUE)
# image(seq(-6,6,0.2), seq(-4,4,0.2), matrix(pred.grid$Comp.1["posterior",][[1]][,1], length(seq(-6,6,0.2))))
# contour(seq(-6,6,0.2), seq(-4,4,0.2), matrix(pred.grid$Comp.1["posterior",][[1]][,1], length(seq(-6,6,0.2))), add = TRUE)
# image(seq(-6,6,0.2), seq(-4,4,0.2), matrix(pred.grid$Comp.2["posterior",][[1]][,1], length(seq(-6,6,0.2))))
# contour(seq(-6,6,0.2), seq(-4,4,0.2), matrix(pred.grid$Comp.2["posterior",][[1]][,1], length(seq(-6,6,0.2))), add = TRUE)
# pred.grid <- predict(res, newdata = grid, local.aggregate = TRUE)
# image(seq(-6,6,0.2), seq(-4,4,0.2), matrix(pred.grid[[1]]$decision, length(seq(-6,6,0.2))))
# contour(seq(-6,6,0.2), seq(-4,4,0.2), matrix(pred.grid[[1]]$decision, length(seq(-6,6,0.2))), add = TRUE)
# points(d$x, col = as.numeric(d$y)+2)
# image(seq(-6,6,0.2), seq(-4,4,0.2), matrix(pred.grid[[1]]$posterior[,1], length(seq(-6,6,0.2))))
# contour(seq(-6,6,0.2), seq(-4,4,0.2), matrix(pred.grid[[1]]$posterior[,1], length(seq(-6,6,0.2))), add = TRUE)
# points(d$x, col = as.numeric(d$y)+2)
# ## plot predicted local membership
# pred.loc.grid <- predict(res@concomitant, newdata = grid)
# image(seq(-6,6,0.2), seq(-4,4,0.2), matrix(pred.loc.grid[,1], length(seq(-6,6,0.2))))
schiffner/locClass documentation built on May 29, 2019, 3:39 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.
context("FLXMCLsvm")
test_that("FLXMCLsvm: misspecified arguments", {
# wrong variable names
cluster <- kmeans(iris[,1:4], centers = 3)$cluster
expect_that(fit <- flexmix(Species ~ V1, data = iris, concomitant = FLXPmultinom(~ Sepal.Length), model = FLXMCLsvm(kernel = "linear", fitted = FALSE), cluster = cluster, control = list(iter.max = 200, classify = "hard")), throws_error("object 'V1' not found"))
expect_that(fit <- flexmix(Species ~ Sepal.Length, data = iris, concomitant = FLXPmultinom(~ V1), model = FLXMCLsvm(kernel = "linear", fitted = FALSE), cluster = cluster, control = list(iter.max = 200, classify = "hard")), throws_error("object 'V1' not found"))
expect_that(fit <- flexmix(y ~ Sepal.Length, data = iris, concomitant = FLXPmultinom(~ Sepal.Width), model = FLXMCLsvm(kernel = "linear", fitted = FALSE), cluster = cluster, control = list(iter.max = 200, classify = "hard")), throws_error("object 'y' not found"))
# wrong class
expect_error(fit <- flexmix(iris, data = iris, concomitant = FLXPmultinom(~ Sepal.Width), model = FLXMCLsvm(kernel = "linear", fitted = FALSE), cluster = cluster, control = list(iter.max = 200, classify = "hard")))
# target variable also in x
# expect_error(mob(Species ~ Species + Sepal.Length | Sepal.Width, data = iris, model = majorityModel,
# control = mob_control(objfun = deviance, minsplit = 20))) ## funktioniert, sollte aber nicht
})
test_that("FLXMCLsvm: arguments are passed to wsvm",{
iris[,1:4] <- scale(iris[,1:4])
set.seed(123)
cluster <- kmeans(iris[,"Sepal.Width"], centers = 3)$cluster
## kernel, cost
fit <- flexmix(Species ~ Sepal.Width, data = iris, model = FLXMCLsvm(kernel = "linear", cost = 2), cluster = cluster, control = list(iter.max = 50, classify = "weighted", verb = 1))
expect_equal(fit@components[[1]][[1]]@parameters$kernel, 0)
expect_equal(fit@components[[1]][[1]]@parameters$cost, 2)
expect_true(fit@components[[1]][[1]]@parameters$fitted)
## degree, coef0
fit <- flexmix(Species ~ Sepal.Width, data = iris, model = FLXMCLsvm(kernel = "polynomial", coef0 = 2, degree = 2, fitted = TRUE, tolerance = 10^-5), cluster = cluster, control = list(iter.max = 50, classify = "weighted", verb = 1))
expect_equal(fit@components[[1]][[1]]@parameters$kernel, 1)
expect_equal(fit@components[[1]][[1]]@parameters$coef0, 2)
expect_equal(fit@components[[1]][[1]]@parameters$degree, 2)
expect_true(fit@components[[1]][[1]]@parameters$fitted)
## gamma
fit <- flexmix(Species ~ Sepal.Width, data = iris, model = FLXMCLsvm(kernel = "radial", gamma = 2, fitted = FALSE, tolerance = 10^-5), cluster = cluster, control = list(iter.max = 50, classify = "weighted", verb = 1))
expect_equal(fit@components[[1]][[1]]@parameters$kernel, 2)
expect_equal(fit@components[[1]][[1]]@parameters$gamma, 2)
expect_true(!fit@components[[1]][[1]]@parameters$fitted)
})
test_that("FLXMCLsvm with several options works",{
cluster <- kmeans(iris[,1], centers = 2)$cluster
# ## weighted, FLXPwlda
# fit <- flexmix(Species ~ Sepal.Width, data = iris, concomitant = FLXPwlda(~ Sepal.Length), model = FLXMCLsvm(kernel = "linear", fitted = FALSE), cluster = cluster, control = list(iter.max = 200, classify = "weighted", verb = 1))
# # ok
# ## hard, FLXPwlda
# fit <- flexmix(Species ~ Sepal.Width, data = iris, concomitant = FLXPwlda(~ Sepal.Length), model = FLXMCLsvm(kernel = "linear", fitted = FALSE), cluster = cluster, control = list(iter.max = 200, classify = "hard", verb = 1))
# # not monotone, empty classes
## weighted, FLXPmultinom
fit <- flexmix(Species ~ Sepal.Width, data = iris, concomitant = FLXPmultinom(~ Sepal.Length), model = FLXMCLsvm(kernel = "linear", fitted = FALSE), cluster = cluster, control = list(iter.max = 200, classify = "weighted", verb = 1))
# ok
## hard, FLXPmultinom
fit <- flexmix(Species ~ Sepal.Width, data = iris, concomitant = FLXPmultinom(~ Sepal.Length), model = FLXMCLsvm(kernel = "linear", fitted = FALSE), cluster = cluster, control = list(iter.max = 200, classify = "hard", verb = 1))
# ok
})
test_that("FLXMCLsvm throws a warning if grouping variable is numeric", {
iris$Species <- as.numeric(iris$Species)
cluster <- kmeans(iris[,c(1,3)], centers = 2)$cluster
expect_that(tr2 <- flexmix(Species ~ Sepal.Length, data = iris, concomitant = FLXPmultinom(~ Petal.Length + Sepal.Length), model = FLXMCLsvm(kernel = "linear", fitted = FALSE, tolerance = 10^-6), cluster = cluster, control = list(iter.max = 50, classify = "weighted", tolerance = 10^-4, verb = 1)), gives_warning("'grouping' was coerced to a factor"))
## use weighted here instead of hard, otherwise SVM does not work
})
test_that("FLXMCLsvm works if only one predictor variable is given", {
cluster <- kmeans(iris[,1:4], centers = 2)$cluster
fit <- flexmix(Species ~ Sepal.Width, data = iris, concomitant = FLXPmultinom(~ Sepal.Length), model = FLXMCLsvm(kernel = "linear", fitted = FALSE), cluster = cluster, control = list(iter.max = 200, classify = "weighted", verb = 1))
})
test_that("FLXMCLsvm: Local and global solution coincide if only one cluster is given", {
fit <- flexmix(Species ~ ., data = iris, model = FLXMCLsvm(kernel = "linear", fitted = FALSE), cluster = 1, control = list(iter.max = 200, classify = "hard"))
w <- wsvm(Species ~ ., data = iris, kernel = "linear")
expect_equal(fit@components[[1]][[1]]@parameters$coefs, w$coefs)
pred <- max.col(mypredict(fit)[[1]])
p <- predict(w)
expect_equal(pred, as.numeric(p))
})
test_that("FLXMCLsvm: training data from only one class", {
cluster <- kmeans(iris[1:50,1:4], centers = 3)$cluster
expect_that(fit <- flexmix(Species ~ Sepal.Width + Sepal.Length, data = iris[1:50,], concomitant = FLXPmultinom(~ Sepal.Width + Sepal.Length), model = FLXMCLsvm(kernel = "linear", fitted = FALSE), cluster = cluster, control = list(iter.max = 200, classify = "hard")), throws_error("need training data from at least two classes"))
})
test_that("FLXMCLsvm: missing classes in individual clusters", {
set.seed(123)
cluster <- kmeans(iris[,1], centers = 2)$cluster
expect_that(fit <- flexmix(Species ~ Sepal.Width, data = iris, concomitant = FLXPmultinom(~ Sepal.Length), model = FLXMCLsvm(kernel = "linear", fitted = FALSE), cluster = cluster, control = list(iter.max = 100, classify = "hard")), gives_warning("some groups are empty"))
expect_equal(ncol(fit@components$Comp.1[[1]]@parameters$coefs), 2)
expect_equal(ncol(fit@components$Comp.2[[1]]@parameters$coefs), 1)
pred <- mypredict(fit, aggregate = FALSE)
expect_equal(colnames(pred$Comp.1), levels(iris$Species))
expect_equal(colnames(pred$Comp.2), levels(iris$Species))
expect_true(all(pred$Comp.2[,1] == 0))
irisna <- iris
irisna$Sepal.Width[1:12] <- NA
pred <- mypredict(fit, newdata = irisna, aggregate = FALSE)
expect_equal(colnames(pred$Comp.1), levels(iris$Species))
expect_equal(colnames(pred$Comp.2), levels(iris$Species))
expect_true(all(pred$Comp.2[-c(1:12),1] == 0))
# library(mlbench)
# data(Glass)
# set.seed(120)
# cluster <- kmeans(Glass[,1], centers = 2)$cluster
# expect_that(fit <- flexmix(Type ~ Na, data = Glass, concomitant = FLXPwlda(as.formula(paste("~", paste(colnames(Glass)[1:9], collapse = "+")))), model = FLXMCLsvm(kernel = "linear", fitted = FALSE), cluster = cluster, control = list(iter.max = 200, classify = "hard", verb = 1)), gives_warning("some groups are empty"))
# expect_equal(ncol(fit@components$Comp.1[[1]]@parameters$coefs), 4)
# expect_equal(ncol(fit@components$Comp.2[[1]]@parameters$coefs), 5)
# pred <- mypredict(fit, aggregate = FALSE)
# expect_equal(colnames(pred$Comp.1), as.character(c(1:3,5:7)))
# expect_equal(colnames(pred$Comp.2), as.character(c(1:3,5:7)))
# expect_equal(colnames(pred$Comp.1[[2]]), c("1/2", "1/3", "1/5", "1/6", "1/7", "2/3", "2/5", "2/6", "2/7", "3/5", "3/6", "3/7", "5/6", "5/7", "6/7"))
# expect_equal(colnames(pred$Comp.2[[2]]), c("1/2", "1/3", "1/5", "1/6", "1/7", "2/3", "2/5", "2/6", "2/7", "3/5", "3/6", "3/7", "5/6", "5/7", "6/7"))
# expect_true(all(pred$Comp.2[[1]][,5] == 0))
# expect_true(all(pred$Comp.2[[2]][,c(4,8,11,13,15)] == 0))
# Glassna <- Glass
# Glassna$Na[1:12] <- NA
# pred <- mypredict(fit, newdata = Glassna, aggregate = FALSE)
# expect_equal(colnames(pred$Comp.1[[1]]), as.character(c(1:3,5:7)))
# expect_equal(colnames(pred$Comp.2[[1]]), as.character(c(1:3,5:7)))
# expect_equal(colnames(pred$Comp.1[[2]]), c("1/2", "1/3", "1/5", "1/6", "1/7", "2/3", "2/5", "2/6", "2/7", "3/5", "3/6", "3/7", "5/6", "5/7", "6/7"))
# expect_equal(colnames(pred$Comp.2[[2]]), c("1/2", "1/3", "1/5", "1/6", "1/7", "2/3", "2/5", "2/6", "2/7", "3/5", "3/6", "3/7", "5/6", "5/7", "6/7"))
# expect_true(all(pred$Comp.2[[1]][-c(1:12),5] == 0))
# expect_true(all(pred$Comp.2[[2]][-c(1:12),c(4,8,11,13,15)] == 0))
# expect_true(all(is.na(pred$Comp.1[[1]][1:12,])))
# expect_true(all(is.na(pred$Comp.1[[2]][1:12,])))
# expect_true(all(is.na(pred$Comp.2[[1]][1:12,])))
# expect_true(all(is.na(pred$Comp.2[[2]][1:12,])))
})
test_that("FLXMCLsvm: removing clusters works", {
set.seed(120)
data <- benchData::flashData(500)
cluster <- kmeans(data$x, centers = 12)$cluster
tr2 <- flexmix(y ~ ., data = as.data.frame(data), concomitant = FLXPmultinom(~ x.1 + x.2), model = FLXMCLsvm(tolerance = 10^-6), cluster = cluster, control = list(iter.max = 100, verb = 1))
expect_equal(length(tr2@components), 9)
expect_equal(ncol(tr2@posterior$scaled), 9)
})
#=================================================================================================================
context("predict FLXMCLsvm")
test_that("predict FLXMCLsvm works correctly with missing newdata", {
set.seed(120)
ran <- sample(1:500,300)
data <- benchData::flashData(500)
cluster <- kmeans(data$x[ran,], centers = 2)$cluster
tr2 <- flexmix(y ~ ., data = as.data.frame(data)[ran,], concomitant = FLXPmultinom(~ x.1 + x.2), model = FLXMCLsvm(kernel = "linear", fitted = FALSE), cluster = cluster, control = list(iter.max = 200))
pred1 <- mypredict(tr2, aggregate = FALSE)
pred2 <- mypredict(tr2, aggregate = FALSE, newdata = as.data.frame(data)[ran,])
expect_equal(pred1, pred2)
expect_equal(rownames(pred1$Comp.1), rownames(as.data.frame(data)[ran,]))
expect_equal(rownames(pred1$Comp.1), rownames(as.data.frame(data)[ran,]))
pred1 <- mypredict(tr2, aggregate = TRUE)
pred2 <- mypredict(tr2, aggregate = TRUE, newdata = as.data.frame(data)[ran,])
expect_equal(pred1, pred2)
expect_equal(rownames(pred1[[1]]), rownames(as.data.frame(data)[ran,]))
})
test_that("predict FLXMCLsvm works with missing classes in the training data", {
ran <- sample(1:150,100)
cluster <- kmeans(iris[1:100,1:4], centers = 2)$cluster
expect_that(tr2 <- flexmix(Species ~ ., data = iris[1:100,], concomitant = FLXPmultinom(~ Sepal.Width + Sepal.Length), model = FLXMCLsvm(), cluster = cluster, control = list(iter.max = 200)), gives_warning("some groups are empty"))
pred <- mypredict(tr2, aggregate = FALSE)
expect_equal(ncol(pred[[1]]), 3) #!!!
expect_equal(ncol(pred[[2]]), 3) #!!!
expect_true(all(pred[[1]][,3] == 0))
expect_true(all(pred[[2]][,3] == 0))
pred <- mypredict(tr2, aggregate = TRUE)
expect_equal(ncol(pred[[1]]), 3) #!!!
expect_true(all(pred[[1]][,3] == 0))
})
test_that("predict FLXMCLsvm works with one single predictor variable", {
ran <- sample(1:150,100)
cluster <- kmeans(iris[ran,1:4], centers = 2)$cluster
tr2 <- flexmix(Species ~ Sepal.Width, data = iris[ran,], concomitant = FLXPmultinom(~ Sepal.Width), model = FLXMCLsvm(kernel = "linear", fitted = FALSE), cluster = cluster, control = list(iter.max = 200))
pred <- mypredict(tr2, newdata = iris[-ran,], aggregate = FALSE)
pred <- mypredict(tr2, aggregate = TRUE)
})
test_that("predict FLXMCLsvm works with one single test observation", {
set.seed(120)
ran <- sample(1:150,100)
cluster <- kmeans(iris[ran,1:4], centers = 2)$cluster
tr2 <- flexmix(Species ~ Sepal.Width, data = iris[ran,], concomitant = FLXPmultinom(~ Sepal.Width), model = FLXMCLsvm(kernel = "linear", fitted = FALSE), cluster = cluster, control = list(iter.max = 200))
pred <- mypredict(tr2, newdata = iris[5,])
expect_equal(dim(pred[[1]]), c(1,3))
expect_equal(dim(pred[[2]]), c(1,3))
pred <- mypredict(tr2, newdata = iris[5,], aggregate = TRUE)
expect_equal(dim(pred[[1]]), c(1,3))
})
test_that("predict FLXMCLsvm: NA handling in newdata works", {
ran <- sample(1:150,100)
cluster <- kmeans(iris[ran,1:4], centers = 2)$cluster
tr2 <- flexmix(Species ~ Sepal.Length + Sepal.Width, data = iris[ran,], concomitant = FLXPmultinom(~ Sepal.Width + Petal.Width), model = FLXMCLsvm(kernel = "linear", fitted = FALSE), cluster = cluster, control = list(iter.max = 200))
## NAs in explanatory variables are ok
irisna <- iris
irisna[1:17,c(1,3)] <- NA
pred <- mypredict(tr2, newdata = irisna)
expect_equal(all(is.na(pred[[1]][1:17,])), TRUE)
expect_equal(all(is.na(pred[[2]][1:17,])), TRUE)
pred <- mypredict(tr2, newdata = irisna, aggregate = TRUE)
expect_equal(all(is.na(pred[[1]][1:17,])), TRUE)
## NAs in splitting variable are not ok if aggregation is desired
irisna[1:17,1:3] <- NA
pred <- mypredict(tr2, newdata = irisna)
expect_equal(all(is.na(pred[[1]][1:17,])), TRUE)
expect_equal(all(is.na(pred[[2]][1:17,])), TRUE)
pred <- mypredict(tr2, newdata = irisna, aggregate = TRUE)
expect_equal(all(is.na(pred[[1]][1:17,])), TRUE)
})
test_that("predict FLXMCLsvm: misspecified arguments", {
ran <- sample(1:150,100)
cluster <- kmeans(iris[ran,1:4], centers = 2)$cluster
tr2 <- flexmix(Species ~ Sepal.Width + Petal.Width, data = iris[ran,], concomitant = FLXPmultinom(~ Sepal.Width + Petal.Width), model = FLXMCLsvm(), cluster = cluster, control = list(iter.max = 200))
# errors in newdata
expect_error(mypredict(tr2, newdata = TRUE))
expect_error(mypredict(tr2, newdata = -50:50))
})
#=================================================================================================================
# library(locClass)
# d <- flashData(500)
# #d <- vNormalData(500)
# grid <- expand.grid(x.1=seq(-6,6,0.2), x.2=seq(-4,4,0.2))
# model <- FLXMCLsvm(kernel = "linear")
# cluster <- kmeans(d$x, center = 2)$cluster
# res <- flexmix(y ~ ., data = as.data.frame(d), concomitant = FLXPmultinom(~ x.1 + x.2), model = model, cluster = cluster)
# res
# plot(d$x, col = res@cluster, cex = res@posterior$scaled[,1])
# plot(d$x, col = res@cluster, cex = res@posterior$scaled[,2])
# plot(d$x, col = d$y, cex = res@posterior$scaled[,1])
# plot(d$x, col = d$y, cex = res@posterior$scaled[,2])
# pred <- predict(res, newdata = as.data.frame(d), local.aggregate = TRUE)
# pred.grid <- predict(res, newdata = grid)
# image(seq(-6,6,0.2), seq(-4,4,0.2), matrix(pred.grid$Comp.1["decision",][[1]], length(seq(-6,6,0.2))))
# contour(seq(-6,6,0.2), seq(-4,4,0.2), matrix(pred.grid$Comp.1["decision",][[1]], length(seq(-6,6,0.2))), add = TRUE)
# image(seq(-6,6,0.2), seq(-4,4,0.2), matrix(pred.grid$Comp.2["decision",][[1]], length(seq(-6,6,0.2))))
# contour(seq(-6,6,0.2), seq(-4,4,0.2), matrix(pred.grid$Comp.2["decision",][[1]], length(seq(-6,6,0.2))), add = TRUE)
# image(seq(-6,6,0.2), seq(-4,4,0.2), matrix(pred.grid$Comp.1["posterior",][[1]][,1], length(seq(-6,6,0.2))))
# contour(seq(-6,6,0.2), seq(-4,4,0.2), matrix(pred.grid$Comp.1["posterior",][[1]][,1], length(seq(-6,6,0.2))), add = TRUE)
# image(seq(-6,6,0.2), seq(-4,4,0.2), matrix(pred.grid$Comp.2["posterior",][[1]][,1], length(seq(-6,6,0.2))))
# contour(seq(-6,6,0.2), seq(-4,4,0.2), matrix(pred.grid$Comp.2["posterior",][[1]][,1], length(seq(-6,6,0.2))), add = TRUE)
# pred.grid <- predict(res, newdata = grid, local.aggregate = TRUE)
# image(seq(-6,6,0.2), seq(-4,4,0.2), matrix(pred.grid[[1]]$decision, length(seq(-6,6,0.2))))
# contour(seq(-6,6,0.2), seq(-4,4,0.2), matrix(pred.grid[[1]]$decision, length(seq(-6,6,0.2))), add = TRUE)
# points(d$x, col = as.numeric(d$y)+2)
# image(seq(-6,6,0.2), seq(-4,4,0.2), matrix(pred.grid[[1]]$posterior[,1], length(seq(-6,6,0.2))))
# contour(seq(-6,6,0.2), seq(-4,4,0.2), matrix(pred.grid[[1]]$posterior[,1], length(seq(-6,6,0.2))), add = TRUE)
# points(d$x, col = as.numeric(d$y)+2)
# ## plot predicted local membership
# pred.loc.grid <- predict(res@concomitant, newdata = grid)
# image(seq(-6,6,0.2), seq(-4,4,0.2), matrix(pred.loc.grid[,1], length(seq(-6,6,0.2))))
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.