tests/testthat/test-vg.R
In AllenInstitute/scrattch.hicat: Hierarchical Iterative Clustering Analysis for Transcriptomic data
context("test-vg")
library(scrattch.hicat)
library(devtools)
# Load glial test data
devtools::install_github("AllenInstitute/tasic2016data")
library(tasic2016data)
glial_classes <- c("Astrocyte", "Endothelial Cell", "Microglia",
"Oligodendrocyte", "Oligodendrocyte Precursor Cell")
glial_cells <- tasic_2016_anno[tasic_2016_anno$broad_type %in% glial_classes, ]
glial_cells <- glial_cells[glial_cells$secondary_type_id == 0, ]
glial_data <- log2(tasic_2016_counts[, glial_cells$sample_name] + 1)
glial_data_sparse <- as(glial_data, "dgCMatrix")
check_genes <- c("Opalin", "Hspa8", "Mbp")
check_idx <- match(check_genes, rownames(glial_data))
## rescale_samples() tests
test_that(
"rescale_samples() rescales data for both matrix and dgCMatrix",
{
rescaled_data <- rescale_samples(glial_data)
rescaled_data_sparse <- rescale_samples(glial_data_sparse)
test_sums <- apply(glial_data, 2, sum)
test_med <- median(test_sums)
test_scales <- test_sums / test_med
test_rescaled <- glial_data / test_scales[col(glial_data)]
test_rescaled_sparse <- as(test_rescaled, "dgCMatrix")
expect_is(rescaled_data, "matrix")
expect_equal(rescaled_data, test_rescaled)
expect_is(rescaled_data_sparse, "dgCMatrix")
expect_equal(rescaled_data_sparse@x, test_rescaled_sparse@x)
expect_equal(rescaled_data, as(rescaled_data_sparse, "matrix"))
}
)
## gene_means() tests
test_that(
"gene_means() computes means for both matrix and dgCMatrix",
{
# no rescaling
mean_values <- gene_means(glial_data,
rescale = FALSE)
mean_values_sparse <- gene_means(glial_data_sparse,
rescale = FALSE)
test_means <- apply(glial_data[check_genes,], 1, mean)
expect_is(mean_values, "numeric")
expect_equal(mean_values[check_idx], test_means)
expect_is(mean_values_sparse, "numeric")
expect_equal(mean_values_sparse[check_idx], test_means)
expect_equal(mean_values, mean_values_sparse)
# with rescaling
mean_values <- gene_means(glial_data,
rescale = TRUE)
mean_values_sparse <- gene_means(glial_data_sparse,
rescale = TRUE)
rescaled_data <- rescale_samples(glial_data)
test_means <- gene_means(rescaled_data[check_idx,],
rescale = FALSE)
expect_is(mean_values, "numeric")
expect_equal(mean_values[check_idx], test_means)
expect_is(mean_values_sparse, "numeric")
expect_equal(mean_values_sparse[check_idx], test_means)
expect_equal(mean_values, mean_values_sparse)
}
)
## gene_vars() tests
test_that(
"gene_vars() computes variances for both matrix and dgCMatrix",
{
# no rescaling, no means
var_values1 <- gene_vars(glial_data,
means = NULL,
rescale = FALSE)
var_values_sparse1 <- gene_vars(glial_data_sparse,
means = NULL,
rescale = FALSE)
test_means <- apply(glial_data[check_genes,], 1, mean)
test_vars <- rowMeans(glial_data[check_genes,] ^ 2) - test_means ^ 2
expect_is(var_values1, "numeric")
expect_equal(var_values1[check_idx], test_vars)
expect_is(var_values_sparse1, "numeric")
expect_equal(var_values_sparse1[check_idx], test_vars)
expect_equal(var_values1, var_values_sparse1)
# no rescaling, provided means
mean_values <- gene_means(glial_data)
var_values2 <- gene_vars(glial_data,
means = mean_values,
rescale = FALSE)
var_values_sparse2 <- gene_vars(glial_data_sparse,
means = mean_values,
rescale = FALSE)
test_means <- apply(glial_data[check_genes,], 1, mean)
test_vars <- rowMeans(glial_data[check_genes,] ^ 2) - test_means ^ 2
expect_is(var_values2, "numeric")
expect_equal(var_values2[check_idx], test_vars)
expect_is(var_values_sparse2, "numeric")
expect_equal(var_values_sparse2[check_idx], test_vars)
expect_equal(var_values2, var_values_sparse2)
# with rescaling, no means
var_values3 <- gene_vars(glial_data,
means = NULL,
rescale = TRUE)
var_values_sparse3 <- gene_vars(glial_data_sparse,
means = NULL,
rescale = TRUE)
glial_data_rescaled <- rescale_samples(glial_data)
test_means <- apply(glial_data_rescaled[check_genes,], 1, mean)
test_vars <- rowMeans(glial_data_rescaled[check_genes,] ^ 2) - test_means ^ 2
expect_is(var_values3, "numeric")
expect_equal(var_values3[check_idx], test_vars)
expect_is(var_values_sparse3, "numeric")
expect_equal(var_values_sparse3[check_idx], test_vars)
expect_equal(var_values3, var_values_sparse3)
# with rescaling, provided means
mean_values <- gene_means(glial_data,
rescale = TRUE)
var_values4 <- gene_vars(glial_data,
means = mean_values,
rescale = TRUE)
var_values_sparse4 <- gene_vars(glial_data_sparse,
means = mean_values,
rescale = TRUE)
glial_data_rescaled <- rescale_samples(glial_data)
test_means <- apply(glial_data_rescaled[check_genes,], 1, mean)
test_vars <- rowMeans(glial_data_rescaled[check_genes,] ^ 2) - test_means ^ 2
expect_is(var_values4, "numeric")
expect_equal(var_values4[check_idx], test_vars)
expect_is(var_values_sparse4, "numeric")
expect_equal(var_values_sparse4[check_idx], test_vars)
expect_equal(var_values4, var_values_sparse4)
# cross conditions
expect_equal(var_values1, var_values2)
expect_equal(var_values3, var_values4)
}
)
## gene_dispersion() tests
test_that(
"gene_dispersion() computes dispersion values using matrix and dgCMatrix inputs",
{
# no means or vars; no rescaling
disp_values1 <- gene_dispersion(glial_data,
means = NULL,
vars = NULL,
rescale = FALSE)
disp_values_sparse1 <- gene_dispersion(glial_data_sparse,
means = NULL,
vars = NULL,
rescale = FALSE)
test_means <- apply(glial_data[check_genes,], 1, mean)
test_vars <- rowMeans(glial_data[check_genes,] ^ 2) - test_means ^ 2
test_disp <- log10(test_vars / test_means)
expect_is(disp_values1, "numeric")
expect_equal(disp_values1[check_idx], test_disp)
expect_is(disp_values_sparse1, "numeric")
expect_equal(disp_values_sparse1[check_idx], test_disp)
expect_equal(disp_values1, disp_values_sparse1)
# no means or vars; with rescaling
disp_values2 <- gene_dispersion(glial_data,
means = NULL,
vars = NULL,
rescale = TRUE)
disp_values_sparse2 <- gene_dispersion(glial_data_sparse,
means = NULL,
vars = NULL,
rescale = TRUE)
glial_data_rescaled <- rescale_samples(glial_data)
test_means <- apply(glial_data_rescaled[check_genes,], 1, mean)
test_vars <- rowMeans(glial_data_rescaled[check_genes,] ^ 2) - test_means ^ 2
test_disp <- log10(test_vars / test_means)
expect_is(disp_values2, "numeric")
expect_equal(disp_values2[check_idx], test_disp)
expect_is(disp_values_sparse2, "numeric")
expect_equal(disp_values_sparse2[check_idx], test_disp)
expect_equal(disp_values2, disp_values_sparse2)
# supplied means and vars; no rescaling
mean_values <- gene_means(glial_data)
var_values <- gene_vars(glial_data,
means = mean_values)
disp_values3 <- gene_dispersion(glial_data,
means = mean_values,
vars = var_values,
rescale = FALSE)
disp_values_sparse3 <- gene_dispersion(glial_data_sparse,
means = mean_values,
vars = var_values,
rescale = FALSE)
test_means <- apply(glial_data[check_genes,], 1, mean)
test_vars <- rowMeans(glial_data[check_genes,] ^ 2) - test_means ^ 2
test_disp <- log10(test_vars / test_means)
expect_is(disp_values3, "numeric")
expect_equal(disp_values3[check_idx], test_disp)
expect_is(disp_values_sparse3, "numeric")
expect_equal(disp_values_sparse3[check_idx], test_disp)
expect_equal(disp_values3, disp_values_sparse3)
# supplied means and vars; with rescaling
mean_values <- gene_means(glial_data,
rescale = TRUE)
var_values <- gene_vars(glial_data,
means = mean_values,
rescale = TRUE)
disp_values4 <- gene_dispersion(glial_data,
means = mean_values,
vars = var_values,
rescale = TRUE)
disp_values_sparse4 <- gene_dispersion(glial_data_sparse,
means = mean_values,
vars = var_values,
rescale = TRUE)
test_means <- apply(glial_data_rescaled[check_genes,], 1, mean)
test_vars <- rowMeans(glial_data_rescaled[check_genes,] ^ 2) - test_means ^ 2
test_disp <- log10(test_vars / test_means)
expect_is(disp_values4, "numeric")
expect_equal(disp_values4[check_idx], test_disp)
expect_is(disp_values_sparse4, "numeric")
expect_equal(disp_values_sparse4[check_idx], test_disp)
expect_equal(disp_values4, disp_values_sparse4)
# cross-condition comparisons
# only conditions with matching rescaling will match
expect_equal(disp_values1, disp_values3)
expect_equal(disp_values2, disp_values4)
}
)
## gene_z() tests
test_that(
"gene_z() returns z values. No tests for accuracy.",
{
# no rescaling, no dispersions
z_values1 <- gene_z(glial_data,
dispersions = NULL,
rescale = FALSE)
z_values_sparse1 <- gene_z(glial_data_sparse,
dispersions = NULL,
rescale = FALSE)
expect_is(z_values1, "numeric")
expect_is(z_values_sparse1, "numeric")
expect_equal(z_values1, z_values_sparse1)
# no rescaling, provided dispersions
disp_values <- gene_dispersion(glial_data)
z_values2 <- gene_z(glial_data,
dispersions = disp_values,
rescale = FALSE)
z_values_sparse2 <- gene_z(glial_data_sparse,
dispersions = disp_values,
rescale = FALSE)
expect_is(z_values2, "numeric")
expect_is(z_values_sparse2, "numeric")
expect_equal(z_values2, z_values_sparse2)
# with rescaling, no dispersions
z_values3 <- gene_z(glial_data,
dispersions = NULL,
rescale = TRUE)
z_values_sparse3 <- gene_z(glial_data_sparse,
dispersions = NULL,
rescale = TRUE)
expect_is(z_values3, "numeric")
expect_is(z_values_sparse3, "numeric")
expect_equal(z_values3, z_values_sparse3)
# with rescaling, provided dispersions
disp_values <- gene_dispersion(glial_data,
rescale = TRUE)
z_values4 <- gene_z(glial_data,
dispersions = disp_values,
rescale = TRUE)
z_values_sparse4 <- gene_z(glial_data_sparse,
dispersions = disp_values,
rescale = TRUE)
expect_is(z_values4, "numeric")
expect_is(z_values_sparse4, "numeric")
expect_equal(z_values4, z_values_sparse4)
# cross condition comparisons
expect_equal(z_values1, z_values2)
expect_equal(z_values3, z_values4)
}
)
## gene_loess_z() tests
test_that(
"gene_loess_fit() computes loess fit for dispersions. No tests for accuracy.",
{
# no rescaling, no dispersions
loess_fit_values1 <- gene_loess_fit(glial_data,
dispersions = NULL,
means = NULL,
rescale = FALSE)
loess_fit_values_sparse1 <- gene_loess_fit(glial_data_sparse,
dispersions = NULL,
means = NULL,
rescale = FALSE)
expect_is(loess_fit_values1, "loess")
expect_is(loess_fit_values_sparse1, "loess")
expect_equal(loess_fit_values1, loess_fit_values_sparse1)
# no rescaling, provided dispersions
disp_values <- gene_dispersion(glial_data)
mean_values <- gene_means(glial_data)
loess_fit_values2 <- gene_loess_fit(glial_data,
dispersions = disp_values,
means = mean_values,
rescale = FALSE)
loess_fit_values_sparse2 <- gene_loess_fit(glial_data_sparse,
dispersions = disp_values,
means = mean_values,
rescale = FALSE)
expect_is(loess_fit_values2, "loess")
expect_is(loess_fit_values_sparse2, "loess")
expect_equal(loess_fit_values2, loess_fit_values_sparse2)
# with rescaling, no dispersions
loess_fit_values3 <- gene_loess_fit(glial_data,
dispersions = NULL,
means = NULL,
rescale = TRUE)
loess_fit_values_sparse3 <- gene_loess_fit(glial_data_sparse,
dispersions = NULL,
means = NULL,
rescale = TRUE)
expect_is(loess_fit_values3, "loess")
expect_is(loess_fit_values_sparse3, "loess")
expect_equal(loess_fit_values3, loess_fit_values_sparse3)
# with rescaling, provided dispersions
mean_values <- gene_means(glial_data,
rescale = TRUE)
disp_values <- gene_dispersion(glial_data,
rescale = TRUE)
loess_fit_values4 <- gene_loess_fit(glial_data,
dispersions = disp_values,
means = mean_values,
rescale = TRUE)
loess_fit_values_sparse4 <- gene_loess_fit(glial_data_sparse,
dispersions = disp_values,
means = mean_values,
rescale = TRUE)
expect_is(loess_fit_values4, "loess")
expect_is(loess_fit_values_sparse4, "loess")
expect_equal(loess_fit_values4, loess_fit_values_sparse4)
# cross condition comparisons
expect_equal(loess_fit_values1, loess_fit_values2)
expect_equal(loess_fit_values3, loess_fit_values4)
}
)
## gene_loess_fit_z() tests
test_that(
"gene_loess_fit_z() computes z values based on a loess fit and dispersion values.",
{
disp_values <- gene_dispersion(glial_data,
means = NULL,
vars = NULL,
rescale = TRUE)
loess_fit <- gene_loess_fit(glial_data,
dispersions = disp_values,
means = NULL,
rescale = TRUE)
loess_z_values <- gene_loess_fit_z(loess_fit,
dispersions = disp_values)
expect_is(loess_z_values, "numeric")
expect_equal(length(loess_z_values), nrow(glial_data))
}
)
## compute_vg() tests
test_that(
"compute_vg_stats() calculates gene variance statistics.",
{
vg_results <- compute_vg_stats(glial_data)
expect_is(vg_results, "data.frame")
sparse_data <- as(glial_data, "dgCMatrix")
sparse_vg_results <- compute_vg_stats(sparse_data)
expect_is(sparse_vg_results, "data.frame")
expect_equal(sparse_vg_results, vg_results)
}
)
## plot_vg() tests
test_that(
"plot_vg() generates plots of gene variance statistics.",
{
glial_stats <- compute_vg_stats(glial_data)
glial_loess <- gene_loess_fit(glial_data)
expect_warning(plot_vg(glial_stats,
plots = "all",
loess_fit = NULL))
all_plots <- suppressWarnings(plot_vg(glial_stats,
plots = "all",
loess_fit = NULL))
expect_is(all_plots, "list")
expect_equal(length(all_plots), 4)
expect_equal(names(all_plots), c("qq_z_plot","qq_loess.z_plot","z_density_plot","loess.z_density_plot"))
expect_is(all_plots$qq_z_plot, "ggplot")
expect_is(all_plots$qq_loess.z_plot, "ggplot")
expect_is(all_plots$z_density_plot, "ggplot")
expect_is(all_plots$loess.z_density_plot, "ggplot")
all_plots <- plot_vg(glial_stats,
plots = "all",
loess_fit = glial_loess)
expect_is(all_plots, "list")
expect_equal(length(all_plots), 5)
expect_equal(names(all_plots), c("qq_z_plot","qq_loess.z_plot","z_density_plot","loess.z_density_plot","dispersion_fit_plot"))
expect_is(all_plots$qq_z_plot, "ggplot")
expect_is(all_plots$qq_loess.z_plot, "ggplot")
expect_is(all_plots$z_density_plot, "ggplot")
expect_is(all_plots$loess.z_density_plot, "ggplot")
expect_is(all_plots$dispersion_fit_plot, "ggplot")
one_plot <- plot_vg(glial_stats,
plots = "qq_z_plot")
expect_is(one_plot, "list")
expect_equal(length(one_plot), 1)
expect_equal(names(one_plot), "qq_z_plot")
expect_is(one_plot$qq_z_plot, "ggplot")
}
)
list_remove <- function(x,
rem) {
list_names <- names(x)
x[list_names != rem]
}
## Remove plotenv - necessary for comparing plots.
strip_plot_env <- function(x) {
if("list" %in% class(x)) {
for(i in 1:length(x)) {
x[[i]] <- list_remove(x[[i]], "plot_env")
}
} else {
x <- list_remove(x, "plot_env")
}
x
}
## find_vg() tests
AllenInstitute/scrattch.hicat documentation built on Oct. 20, 2023, 6:55 a.m.
R Package Documentation
Browse R Packages
We want your feedback!
Note that we can't provide technical support on individual packages. You should contact the package authors for that.
context("test-vg")
library(scrattch.hicat)
library(devtools)
# Load glial test data
devtools::install_github("AllenInstitute/tasic2016data")
library(tasic2016data)
glial_classes <- c("Astrocyte", "Endothelial Cell", "Microglia",
"Oligodendrocyte", "Oligodendrocyte Precursor Cell")
glial_cells <- tasic_2016_anno[tasic_2016_anno$broad_type %in% glial_classes, ]
glial_cells <- glial_cells[glial_cells$secondary_type_id == 0, ]
glial_data <- log2(tasic_2016_counts[, glial_cells$sample_name] + 1)
glial_data_sparse <- as(glial_data, "dgCMatrix")
check_genes <- c("Opalin", "Hspa8", "Mbp")
check_idx <- match(check_genes, rownames(glial_data))
## rescale_samples() tests
test_that(
"rescale_samples() rescales data for both matrix and dgCMatrix",
{
rescaled_data <- rescale_samples(glial_data)
rescaled_data_sparse <- rescale_samples(glial_data_sparse)
test_sums <- apply(glial_data, 2, sum)
test_med <- median(test_sums)
test_scales <- test_sums / test_med
test_rescaled <- glial_data / test_scales[col(glial_data)]
test_rescaled_sparse <- as(test_rescaled, "dgCMatrix")
expect_is(rescaled_data, "matrix")
expect_equal(rescaled_data, test_rescaled)
expect_is(rescaled_data_sparse, "dgCMatrix")
expect_equal(rescaled_data_sparse@x, test_rescaled_sparse@x)
expect_equal(rescaled_data, as(rescaled_data_sparse, "matrix"))
}
)
## gene_means() tests
test_that(
"gene_means() computes means for both matrix and dgCMatrix",
{
# no rescaling
mean_values <- gene_means(glial_data,
rescale = FALSE)
mean_values_sparse <- gene_means(glial_data_sparse,
rescale = FALSE)
test_means <- apply(glial_data[check_genes,], 1, mean)
expect_is(mean_values, "numeric")
expect_equal(mean_values[check_idx], test_means)
expect_is(mean_values_sparse, "numeric")
expect_equal(mean_values_sparse[check_idx], test_means)
expect_equal(mean_values, mean_values_sparse)
# with rescaling
mean_values <- gene_means(glial_data,
rescale = TRUE)
mean_values_sparse <- gene_means(glial_data_sparse,
rescale = TRUE)
rescaled_data <- rescale_samples(glial_data)
test_means <- gene_means(rescaled_data[check_idx,],
rescale = FALSE)
expect_is(mean_values, "numeric")
expect_equal(mean_values[check_idx], test_means)
expect_is(mean_values_sparse, "numeric")
expect_equal(mean_values_sparse[check_idx], test_means)
expect_equal(mean_values, mean_values_sparse)
}
)
## gene_vars() tests
test_that(
"gene_vars() computes variances for both matrix and dgCMatrix",
{
# no rescaling, no means
var_values1 <- gene_vars(glial_data,
means = NULL,
rescale = FALSE)
var_values_sparse1 <- gene_vars(glial_data_sparse,
means = NULL,
rescale = FALSE)
test_means <- apply(glial_data[check_genes,], 1, mean)
test_vars <- rowMeans(glial_data[check_genes,] ^ 2) - test_means ^ 2
expect_is(var_values1, "numeric")
expect_equal(var_values1[check_idx], test_vars)
expect_is(var_values_sparse1, "numeric")
expect_equal(var_values_sparse1[check_idx], test_vars)
expect_equal(var_values1, var_values_sparse1)
# no rescaling, provided means
mean_values <- gene_means(glial_data)
var_values2 <- gene_vars(glial_data,
means = mean_values,
rescale = FALSE)
var_values_sparse2 <- gene_vars(glial_data_sparse,
means = mean_values,
rescale = FALSE)
test_means <- apply(glial_data[check_genes,], 1, mean)
test_vars <- rowMeans(glial_data[check_genes,] ^ 2) - test_means ^ 2
expect_is(var_values2, "numeric")
expect_equal(var_values2[check_idx], test_vars)
expect_is(var_values_sparse2, "numeric")
expect_equal(var_values_sparse2[check_idx], test_vars)
expect_equal(var_values2, var_values_sparse2)
# with rescaling, no means
var_values3 <- gene_vars(glial_data,
means = NULL,
rescale = TRUE)
var_values_sparse3 <- gene_vars(glial_data_sparse,
means = NULL,
rescale = TRUE)
glial_data_rescaled <- rescale_samples(glial_data)
test_means <- apply(glial_data_rescaled[check_genes,], 1, mean)
test_vars <- rowMeans(glial_data_rescaled[check_genes,] ^ 2) - test_means ^ 2
expect_is(var_values3, "numeric")
expect_equal(var_values3[check_idx], test_vars)
expect_is(var_values_sparse3, "numeric")
expect_equal(var_values_sparse3[check_idx], test_vars)
expect_equal(var_values3, var_values_sparse3)
# with rescaling, provided means
mean_values <- gene_means(glial_data,
rescale = TRUE)
var_values4 <- gene_vars(glial_data,
means = mean_values,
rescale = TRUE)
var_values_sparse4 <- gene_vars(glial_data_sparse,
means = mean_values,
rescale = TRUE)
glial_data_rescaled <- rescale_samples(glial_data)
test_means <- apply(glial_data_rescaled[check_genes,], 1, mean)
test_vars <- rowMeans(glial_data_rescaled[check_genes,] ^ 2) - test_means ^ 2
expect_is(var_values4, "numeric")
expect_equal(var_values4[check_idx], test_vars)
expect_is(var_values_sparse4, "numeric")
expect_equal(var_values_sparse4[check_idx], test_vars)
expect_equal(var_values4, var_values_sparse4)
# cross conditions
expect_equal(var_values1, var_values2)
expect_equal(var_values3, var_values4)
}
)
## gene_dispersion() tests
test_that(
"gene_dispersion() computes dispersion values using matrix and dgCMatrix inputs",
{
# no means or vars; no rescaling
disp_values1 <- gene_dispersion(glial_data,
means = NULL,
vars = NULL,
rescale = FALSE)
disp_values_sparse1 <- gene_dispersion(glial_data_sparse,
means = NULL,
vars = NULL,
rescale = FALSE)
test_means <- apply(glial_data[check_genes,], 1, mean)
test_vars <- rowMeans(glial_data[check_genes,] ^ 2) - test_means ^ 2
test_disp <- log10(test_vars / test_means)
expect_is(disp_values1, "numeric")
expect_equal(disp_values1[check_idx], test_disp)
expect_is(disp_values_sparse1, "numeric")
expect_equal(disp_values_sparse1[check_idx], test_disp)
expect_equal(disp_values1, disp_values_sparse1)
# no means or vars; with rescaling
disp_values2 <- gene_dispersion(glial_data,
means = NULL,
vars = NULL,
rescale = TRUE)
disp_values_sparse2 <- gene_dispersion(glial_data_sparse,
means = NULL,
vars = NULL,
rescale = TRUE)
glial_data_rescaled <- rescale_samples(glial_data)
test_means <- apply(glial_data_rescaled[check_genes,], 1, mean)
test_vars <- rowMeans(glial_data_rescaled[check_genes,] ^ 2) - test_means ^ 2
test_disp <- log10(test_vars / test_means)
expect_is(disp_values2, "numeric")
expect_equal(disp_values2[check_idx], test_disp)
expect_is(disp_values_sparse2, "numeric")
expect_equal(disp_values_sparse2[check_idx], test_disp)
expect_equal(disp_values2, disp_values_sparse2)
# supplied means and vars; no rescaling
mean_values <- gene_means(glial_data)
var_values <- gene_vars(glial_data,
means = mean_values)
disp_values3 <- gene_dispersion(glial_data,
means = mean_values,
vars = var_values,
rescale = FALSE)
disp_values_sparse3 <- gene_dispersion(glial_data_sparse,
means = mean_values,
vars = var_values,
rescale = FALSE)
test_means <- apply(glial_data[check_genes,], 1, mean)
test_vars <- rowMeans(glial_data[check_genes,] ^ 2) - test_means ^ 2
test_disp <- log10(test_vars / test_means)
expect_is(disp_values3, "numeric")
expect_equal(disp_values3[check_idx], test_disp)
expect_is(disp_values_sparse3, "numeric")
expect_equal(disp_values_sparse3[check_idx], test_disp)
expect_equal(disp_values3, disp_values_sparse3)
# supplied means and vars; with rescaling
mean_values <- gene_means(glial_data,
rescale = TRUE)
var_values <- gene_vars(glial_data,
means = mean_values,
rescale = TRUE)
disp_values4 <- gene_dispersion(glial_data,
means = mean_values,
vars = var_values,
rescale = TRUE)
disp_values_sparse4 <- gene_dispersion(glial_data_sparse,
means = mean_values,
vars = var_values,
rescale = TRUE)
test_means <- apply(glial_data_rescaled[check_genes,], 1, mean)
test_vars <- rowMeans(glial_data_rescaled[check_genes,] ^ 2) - test_means ^ 2
test_disp <- log10(test_vars / test_means)
expect_is(disp_values4, "numeric")
expect_equal(disp_values4[check_idx], test_disp)
expect_is(disp_values_sparse4, "numeric")
expect_equal(disp_values_sparse4[check_idx], test_disp)
expect_equal(disp_values4, disp_values_sparse4)
# cross-condition comparisons
# only conditions with matching rescaling will match
expect_equal(disp_values1, disp_values3)
expect_equal(disp_values2, disp_values4)
}
)
## gene_z() tests
test_that(
"gene_z() returns z values. No tests for accuracy.",
{
# no rescaling, no dispersions
z_values1 <- gene_z(glial_data,
dispersions = NULL,
rescale = FALSE)
z_values_sparse1 <- gene_z(glial_data_sparse,
dispersions = NULL,
rescale = FALSE)
expect_is(z_values1, "numeric")
expect_is(z_values_sparse1, "numeric")
expect_equal(z_values1, z_values_sparse1)
# no rescaling, provided dispersions
disp_values <- gene_dispersion(glial_data)
z_values2 <- gene_z(glial_data,
dispersions = disp_values,
rescale = FALSE)
z_values_sparse2 <- gene_z(glial_data_sparse,
dispersions = disp_values,
rescale = FALSE)
expect_is(z_values2, "numeric")
expect_is(z_values_sparse2, "numeric")
expect_equal(z_values2, z_values_sparse2)
# with rescaling, no dispersions
z_values3 <- gene_z(glial_data,
dispersions = NULL,
rescale = TRUE)
z_values_sparse3 <- gene_z(glial_data_sparse,
dispersions = NULL,
rescale = TRUE)
expect_is(z_values3, "numeric")
expect_is(z_values_sparse3, "numeric")
expect_equal(z_values3, z_values_sparse3)
# with rescaling, provided dispersions
disp_values <- gene_dispersion(glial_data,
rescale = TRUE)
z_values4 <- gene_z(glial_data,
dispersions = disp_values,
rescale = TRUE)
z_values_sparse4 <- gene_z(glial_data_sparse,
dispersions = disp_values,
rescale = TRUE)
expect_is(z_values4, "numeric")
expect_is(z_values_sparse4, "numeric")
expect_equal(z_values4, z_values_sparse4)
# cross condition comparisons
expect_equal(z_values1, z_values2)
expect_equal(z_values3, z_values4)
}
)
## gene_loess_z() tests
test_that(
"gene_loess_fit() computes loess fit for dispersions. No tests for accuracy.",
{
# no rescaling, no dispersions
loess_fit_values1 <- gene_loess_fit(glial_data,
dispersions = NULL,
means = NULL,
rescale = FALSE)
loess_fit_values_sparse1 <- gene_loess_fit(glial_data_sparse,
dispersions = NULL,
means = NULL,
rescale = FALSE)
expect_is(loess_fit_values1, "loess")
expect_is(loess_fit_values_sparse1, "loess")
expect_equal(loess_fit_values1, loess_fit_values_sparse1)
# no rescaling, provided dispersions
disp_values <- gene_dispersion(glial_data)
mean_values <- gene_means(glial_data)
loess_fit_values2 <- gene_loess_fit(glial_data,
dispersions = disp_values,
means = mean_values,
rescale = FALSE)
loess_fit_values_sparse2 <- gene_loess_fit(glial_data_sparse,
dispersions = disp_values,
means = mean_values,
rescale = FALSE)
expect_is(loess_fit_values2, "loess")
expect_is(loess_fit_values_sparse2, "loess")
expect_equal(loess_fit_values2, loess_fit_values_sparse2)
# with rescaling, no dispersions
loess_fit_values3 <- gene_loess_fit(glial_data,
dispersions = NULL,
means = NULL,
rescale = TRUE)
loess_fit_values_sparse3 <- gene_loess_fit(glial_data_sparse,
dispersions = NULL,
means = NULL,
rescale = TRUE)
expect_is(loess_fit_values3, "loess")
expect_is(loess_fit_values_sparse3, "loess")
expect_equal(loess_fit_values3, loess_fit_values_sparse3)
# with rescaling, provided dispersions
mean_values <- gene_means(glial_data,
rescale = TRUE)
disp_values <- gene_dispersion(glial_data,
rescale = TRUE)
loess_fit_values4 <- gene_loess_fit(glial_data,
dispersions = disp_values,
means = mean_values,
rescale = TRUE)
loess_fit_values_sparse4 <- gene_loess_fit(glial_data_sparse,
dispersions = disp_values,
means = mean_values,
rescale = TRUE)
expect_is(loess_fit_values4, "loess")
expect_is(loess_fit_values_sparse4, "loess")
expect_equal(loess_fit_values4, loess_fit_values_sparse4)
# cross condition comparisons
expect_equal(loess_fit_values1, loess_fit_values2)
expect_equal(loess_fit_values3, loess_fit_values4)
}
)
## gene_loess_fit_z() tests
test_that(
"gene_loess_fit_z() computes z values based on a loess fit and dispersion values.",
{
disp_values <- gene_dispersion(glial_data,
means = NULL,
vars = NULL,
rescale = TRUE)
loess_fit <- gene_loess_fit(glial_data,
dispersions = disp_values,
means = NULL,
rescale = TRUE)
loess_z_values <- gene_loess_fit_z(loess_fit,
dispersions = disp_values)
expect_is(loess_z_values, "numeric")
expect_equal(length(loess_z_values), nrow(glial_data))
}
)
## compute_vg() tests
test_that(
"compute_vg_stats() calculates gene variance statistics.",
{
vg_results <- compute_vg_stats(glial_data)
expect_is(vg_results, "data.frame")
sparse_data <- as(glial_data, "dgCMatrix")
sparse_vg_results <- compute_vg_stats(sparse_data)
expect_is(sparse_vg_results, "data.frame")
expect_equal(sparse_vg_results, vg_results)
}
)
## plot_vg() tests
test_that(
"plot_vg() generates plots of gene variance statistics.",
{
glial_stats <- compute_vg_stats(glial_data)
glial_loess <- gene_loess_fit(glial_data)
expect_warning(plot_vg(glial_stats,
plots = "all",
loess_fit = NULL))
all_plots <- suppressWarnings(plot_vg(glial_stats,
plots = "all",
loess_fit = NULL))
expect_is(all_plots, "list")
expect_equal(length(all_plots), 4)
expect_equal(names(all_plots), c("qq_z_plot","qq_loess.z_plot","z_density_plot","loess.z_density_plot"))
expect_is(all_plots$qq_z_plot, "ggplot")
expect_is(all_plots$qq_loess.z_plot, "ggplot")
expect_is(all_plots$z_density_plot, "ggplot")
expect_is(all_plots$loess.z_density_plot, "ggplot")
all_plots <- plot_vg(glial_stats,
plots = "all",
loess_fit = glial_loess)
expect_is(all_plots, "list")
expect_equal(length(all_plots), 5)
expect_equal(names(all_plots), c("qq_z_plot","qq_loess.z_plot","z_density_plot","loess.z_density_plot","dispersion_fit_plot"))
expect_is(all_plots$qq_z_plot, "ggplot")
expect_is(all_plots$qq_loess.z_plot, "ggplot")
expect_is(all_plots$z_density_plot, "ggplot")
expect_is(all_plots$loess.z_density_plot, "ggplot")
expect_is(all_plots$dispersion_fit_plot, "ggplot")
one_plot <- plot_vg(glial_stats,
plots = "qq_z_plot")
expect_is(one_plot, "list")
expect_equal(length(one_plot), 1)
expect_equal(names(one_plot), "qq_z_plot")
expect_is(one_plot$qq_z_plot, "ggplot")
}
)
list_remove <- function(x,
rem) {
list_names <- names(x)
x[list_names != rem]
}
## Remove plotenv - necessary for comparing plots.
strip_plot_env <- function(x) {
if("list" %in% class(x)) {
for(i in 1:length(x)) {
x[[i]] <- list_remove(x[[i]], "plot_env")
}
} else {
x <- list_remove(x, "plot_env")
}
x
}
## find_vg() tests
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.