tests/testthat/tests-archive/test_gen-summary.R

In AxelitoMartin/gnomeR: Wrangle and analyze IMPACT and TCGA mutation data

# context("check make gen summary")
# source_test_helpers(path = "tests/testthat/helper_dat.R", env = test_env())
#
# test_that("wrong filter",{
#
#   mat = matrix(rnorm(1,100), ncol=4)
#   colnames(mat) = c("Hugo_Symbol", "Variant_Classification", "A","B")
#   outcome <- sample(c("A","B"),nrow(mat),replace = T)
#   expect_error(gen_summary(gen_dat = mat,
#                        outcome = outcome,
#                        filter = -1))
#   expect_error(gen_summary(gen_dat = mat,
#                        outcome = outcome,
#                        filter = 1))
#   expect_error(gen_summary(gen_dat = mat,
#                        outcome = outcome,
#                        filter = 2))
#
# })
#
# #
# # test_that("working binary example",{
# #
# #   outcome <- sample(c("A","B"),nrow(gen_dat),replace = T)
# #   test <- gen_summary(gen_dat = gen_dat,
# #   outcome = outcome,
# #   filter = 0.05,
# #   cont = FALSE,rank = TRUE)
# #
# #   expect_true(is.data.frame(test$fits))
# #   expect_true(ncol(test$fits) == 9)
# #   expect_true(is.ggplot(test$forest.plot))
# #   # expect_true(is.ggplot(test$vPlot))
# # })
#
#
# test_that("working continuous example",{
#
#   gen_dat.test <- gen_dat
#   gen_dat.test[,1] <- rnorm(n = nrow(gen_dat.test))
#   outcome <-  rnorm(n = nrow(gen_dat.test))
#   test <- gen_summary(gen_dat = gen_dat.test,
#                      outcome = outcome,
#                      filter = 0.05,
#                      cont = TRUE,rank = TRUE)
#
#
#   expect_true(is.data.frame(test$fits))
#   expect_true(ncol(test$fits) == 8)
#   expect_true(is.null(test$forest.plot))
#   # expect_true(!is.null(test$vPlot))
# })
#
#
# test_that("working binary example with a feature that is only 1's",{
#
#   gen_dat.test<- gen_dat
#   gen_dat.test[,1] <- 1
#   outcome <-  rnorm(n = nrow(gen_dat.test))
#   test <- gen_summary(gen_dat = gen_dat.test,
#                   outcome = outcome,
#                   filter = 0.05,
#                   cont = TRUE,rank = TRUE)
#
#   expect_true(is.data.frame(test$fits))
#   expect_true(ncol(test$fits) == 8)
#   expect_true(all(class(test$vPlot) == c("plotly","htmlwidget")))
# })
#
#
# test_that("filter too large",{
#
#   gen_dat.test<-gen_dat
#   expect_error(gen_summary(gen_dat = gen_dat.test,
#                   outcome = outcome,
#                   filter = 0.99,
#                   cont = FALSE,rank = TRUE))
#
# })
#
# # test_that("continuous features",{
# #
# #   gen_dat.test<-gen_dat
# #   gen_dat.test[,1] <- rnorm(n = nrow(gen_dat.test))
# #   test <- gen_summary(gen_dat = gen_dat.test,
# #           outcome = outcome,
# #           filter = 0,
# #           cont = FALSE,rank = TRUE)
# #   expect_true(is.data.frame(test$fits))
# #   expect_true(ncol(test$fits) == 9)
# #   expect_true(is.ggplot(test$forest.plot))
# #   # expect_true(is.ggplot(test$vPlot))
# #
# # })
#
#
# # test_that("paired test",{
# #
# #   set.seed(123)
# #   gen_dat <- as.data.frame(matrix(rbinom(500,1,1/2),nrow = 100, ncol = 5))
# #   outcome <- c(rep("Time1",50),rep("Time2",50))
# #   test <- gen_summary(gen_dat = gen_dat,
# #                   outcome = outcome,
# #                   filter = 0,paired = TRUE,
# #                   cont = FALSE,rank = TRUE)
# #   expect_true(is.data.frame(test$fits))
# #   expect_true(ncol(test$fits) == 9)
# #   expect_true(is.ggplot(test$forest.plot))
# #   # expect_true(is.ggplot(test$vPlot))
# #
# # })
#
#
# test_that("three level outcome",{
#
#   set.seed(123)
#   gen_dat <- as.data.frame(matrix(rbinom(500,1,1/2),nrow = 100, ncol = 5))
#   outcome <- sample(c("A","B","C"),100,replace = TRUE)
#   test <- gen_summary(gen_dat = gen_dat,
#                   outcome = outcome,
#                   filter = 0,
#                   cont = FALSE,rank = TRUE)
#   expect_true(is.data.frame(test$fits))
#   expect_true(ncol(test$fits) == 10)
#   expect_true(is.null(test$forest.plot))
#   expect_true(is.null(test$vPlot))
#
# })
#
# test_that("factors hidden in continuous variables",{
#
#   gen_dat.test <- gen_dat
#   gen_dat.test[,1] <- factor(sample(c("NEUTRAL","DELETION","LOH"),nrow(gen_dat.test),replace = TRUE),
#                              levels = c("NEUTRAL","LOH","DELETION"))
#   outcome <-  rnorm(n = nrow(gen_dat.test))
#   test <- gen_summary(gen_dat = gen_dat.test[,-1],
#                   outcome = outcome,
#                   filter = 0,
#                   cont = TRUE,rank = TRUE)
#   expect_true(is.data.frame(test$fits))
#   expect_true(ncol(test$fits) == 8)
#   expect_true(is.null(test$forest.plot))
#   expect_true(!is.null(test$vPlot))
#
# })
#