Rfoehnix: Objective Foehn Diagnosis

library("foehnix")
library("testthat")


# Prepare the models in the "global env" such that we only have
# to run the model estimation twice for all kinds of plots!
# Loading test data
data <- demodata("tyrol")
filter <- list(dd = c(43, 223), crest_dd = c(90, 270))

# Model: model will be added to the global 
# environment such that we can use it in the
# other test_that calls.
mod0 <- foehnix(ff ~ 1, data = data, switch = TRUE,
                filter = filter, verbose = FALSE)
mod1 <- foehnix(diff_t ~ ff + rh, data = data, switch = TRUE,
                filter = filter, verbose = FALSE)

# ------------------------------------------------------------
# --------- testing the foehnix plot function ----------------
# ------------------------------------------------------------
test_that("Testing plot.foehnix function", {

    expect_is(mod0, "foehnix")
    expect_is(mod1, "foehnix")

    # Test plotting functions
    # function(x, which = NULL, log = TRUE, ...) 
    plotfun <- function(x, ...) {
        pdf(file = tempfile())
        plot(x, ...)
        dev.off()
    }

    # Wrong usage
    expect_error(plotfun(mod0, which = 9))
    expect_error(plotfun(mod0, which = c(1,8)))
    expect_error(plotfun(mod0, which = list(1, 2)))
    expect_error(plotfun(mod0, which = "foo"))

    # Testing the different plot types for both models
    expect_silent(plotfun(mod0, ask = FALSE))
    expect_silent(plotfun(mod0, which = "loglik", ask = FALSE))
    expect_silent(plotfun(mod1, which = "loglik", ask = FALSE))
    expect_silent(plotfun(mod0, which = "loglikcontribution", ask = FALSE))
    expect_silent(plotfun(mod1, which = "loglikcontribution", ask = FALSE))
    expect_silent(plotfun(mod0, which = "coef", ask = FALSE))
    expect_silent(plotfun(mod1, which = "coef", ask = FALSE))
    expect_silent(plotfun(mod0, which = "hist", ask = FALSE))
    expect_silent(plotfun(mod1, which = "hist", ask = FALSE))
    expect_silent(plotfun(mod0, which = c("hist", "loglik"), ask = FALSE))
    expect_silent(plotfun(mod1, which = c("hist", "loglik"), ask = FALSE))

    # Integers instead of plot type names
    expect_silent(plotfun(mod0, which = 1:2, ask = FALSE))
    expect_silent(plotfun(mod1, which = 1:2, ask = FALSE))

    # Posterior plot with custom freq/breaks
    expect_silent(plotfun(mod1, which = "post")) 
    expect_silent(plotfun(mod1, which = "post", freq = TRUE))
    expect_silent(plotfun(mod1, which = "post", freq = FALSE))
    expect_error(plotfun(mod1, which = "post", freq = "wrong"))
    expect_silent(plot(mod1, which = "post", breaks = seq(0, 1, by = 0.05)))
    expect_silent(plot(mod1, which = "post", breaks = c(0, 0.2, 0.4, 0.9, 1.0)))

    # Setting log to FALSE
    expect_silent(plotfun(mod0, log = FALSE, ask = FALSE))
    expect_silent(plotfun(mod1, log = FALSE, ask = FALSE))

})



# ------------------------------------------------------------
# ------- testing the image plot function (Hovmoeller) -------
# ------------------------------------------------------------
test_that("Testing image.foehnix function", {

    expect_is(mod1, "foehnix")

    imagefun <- function(x, ...) {
        pdf(file = tempfile())
        res <- image(x, ...)
        dev.off()
        return(res)
    }

    # Test some errors (wrong usage)
    expect_error(foehnix::image.foehnix(data))
    # Wrong function argument
    expect_error(imagefun(mod1, FUN = 3))  
    expect_error(imagefun(mod1, FUN = "foo"))
    # Wrong deltat
    expect_error(imagefun(mod1, deltat = "foo"))
    expect_error(imagefun(mod1, deltat = -3))
    expect_error(imagefun(mod1, deltat = -3))
    expect_error(imagefun(mod1, deltat = 77777))
    # Wrong deltad
    expect_error(imagefun(mod1, deltad = "foo"))
    expect_error(imagefun(mod1, deltad = -3))
    expect_error(imagefun(mod1, deltad = -3))
    expect_error(imagefun(mod1, deltad = 77777))
    # Wrong colors
    expect_error(imagefun(mod1, col = list()))
    expect_error(imagefun(mod1, col = "#ff00ff"))

    expect_silent(res <- imagefun(mod1))
    expect_is(res, "list")
    expect_is(res$agg, "data.frame")
    expect_equal((length(res$breaks.time)-1) * (length(res$breaks.date)-1), nrow(res$agg))

    expect_silent(res <- imagefun(mod1, deltad = 31))
    expect_is(res, "list")
    expect_is(res$agg, "data.frame")
    expect_equal((length(res$breaks.time)-1) * (length(res$breaks.date)-1), nrow(res$agg))

    expect_silent(res <- imagefun(mod1, deltat = 3*3600))
    expect_is(res, "list")
    expect_is(res$agg, "data.frame")
    expect_equal((length(res$breaks.time)-1) * (length(res$breaks.date)-1), nrow(res$agg))

    expect_silent(res <- imagefun(mod1, deltat = 3*3600, deltad = 31))
    expect_is(res, "list")
    expect_is(res$agg, "data.frame")
    expect_equal((length(res$breaks.time)-1) * (length(res$breaks.date)-1), nrow(res$agg))

    # Test if our default functions are handled correctly
    expect_is(imagefun(mod1, FUN = "mean"), "list")
    expect_is(imagefun(mod1, FUN = "occ"), "list")
    expect_is(imagefun(mod1, FUN = "noocc"), "list")
    expect_is(imagefun(mod1, FUN = "freq"), "list")

    # Custom colors
    expect_is(imagefun(mod1, col = rainbow(100)), "list")
    expect_is(imagefun(mod1, contours = TRUE), "list")
    expect_is(imagefun(mod1, contours = TRUE, contour.col = "black"), "list")

    # Custom zlim
    expect_is(res <- imagefun(mod1), "list")
    expect_true(is.null(res$zlim))
    expect_is(res <- imagefun(mod1, zlim = c(0,1)), "list")
    expect_identical(res$zlim, c(0,1))

})