test: create tests

README.md

@@ -2,5 +2,7 @@
 
 Code Ocean capsule: MOSuite - filter low counts
 
+[![tests](https://github.com/CCBR/MOSuite-filter-counts/actions/workflows/tests.yml/badge.svg)](https://github.com/CCBR/MOSuite-filter-counts/actions/workflows/tests.yml)
+
 - [Code Ocean Capsule](https://poc-nci.codeocean.io/capsule/2922767/tree)
 - [MOSuite R package docs](https://ccbr.github.io/MOSuite/)

code/main.R

@@ -1,5 +1,4 @@
 #!/usr/bin/env Rscript
-rlang::global_entrace()
 library(argparse)
 library(glue)
 library(MOSuite)
@@ -13,38 +12,93 @@ setup_capsule_environment()
 # parse CLI arguments
 parser <- ArgumentParser()
 
-parser$add_argument("--count_type", type="character", default="clean")
-parser$add_argument("--feature_id_colname", type="character", default=NULL, help="Column name for feature IDs")
-parser$add_argument("--sample_id_colname", type="character", default=NULL, help="Column name for sample IDs")
-parser$add_argument("--group_colname", type="character", default="Group", help="Column name for sample groups")
-parser$add_argument("--label_colname", type="character", default=NULL, help="Column name for sample labels")
-parser$add_argument("--minimum_count_value_to_be_considered_nonzero", type="integer", default=8, help="Minimum count threshold")
-parser$add_argument("--minimum_number_of_samples_with_nonzero_counts_in_total", type="integer", default=7, help="Minimum samples with nonzero counts")
-parser$add_argument("--minimum_number_of_samples_with_nonzero_counts_in_a_group", type="integer", default=3, help="Minimum samples per group with nonzero counts")
-parser$add_argument("--use_cpm_counts_to_filter", type="logical", default=TRUE, help="Transform to CPM before filtering")
-parser$add_argument("--use_group_based_filtering", type="logical", default=FALSE, help="Use group-based filtering")
-parser$add_argument("--plot_corr_matrix_heatmap", type="logical", default=TRUE, help="Plot correlation heatmap")
-parser$add_argument("--interactive_plots", type="logical", default=FALSE, help="Create interactive plots with plotly")
+parser$add_argument("--count_type", type = "character", default = "clean")
+parser$add_argument(
+  "--feature_id_colname",
+  type = "character",
+  default = NULL,
+  help = "Column name for feature IDs"
+)
+parser$add_argument(
+  "--sample_id_colname",
+  type = "character",
+  default = NULL,
+  help = "Column name for sample IDs"
+)
+parser$add_argument(
+  "--group_colname",
+  type = "character",
+  default = "Group",
+  help = "Column name for sample groups"
+)
+parser$add_argument(
+  "--label_colname",
+  type = "character",
+  default = NULL,
+  help = "Column name for sample labels"
+)
+parser$add_argument(
+  "--minimum_count_value_to_be_considered_nonzero",
+  type = "integer",
+  default = 8,
+  help = "Minimum count threshold"
+)
+parser$add_argument(
+  "--minimum_number_of_samples_with_nonzero_counts_in_total",
+  type = "integer",
+  default = 7,
+  help = "Minimum samples with nonzero counts"
+)
+parser$add_argument(
+  "--minimum_number_of_samples_with_nonzero_counts_in_a_group",
+  type = "integer",
+  default = 3,
+  help = "Minimum samples per group with nonzero counts"
+)
+parser$add_argument(
+  "--use_cpm_counts_to_filter",
+  type = "logical",
+  default = TRUE,
+  help = "Transform to CPM before filtering"
+)
+parser$add_argument(
+  "--use_group_based_filtering",
+  type = "logical",
+  default = FALSE,
+  help = "Use group-based filtering"
+)
+parser$add_argument(
+  "--plot_corr_matrix_heatmap",
+  type = "logical",
+  default = TRUE,
+  help = "Plot correlation heatmap"
+)
+parser$add_argument(
+  "--interactive_plots",
+  type = "logical",
+  default = FALSE,
+  help = "Create interactive plots with plotly"
+)
 
 args <- parser$parse_args()
 
 # load multiOmicDataSet from data directory
 moo <- load_moo_from_data_dir()
 
 # run MOSuite
-moo |> 
-    filter_counts(
-        count_type = args$count_type,
-        feature_id_colname = args$feature_id_colname,
-        sample_id_colname = args$sample_id_colname,
-        group_colname = args$group_colname,
-        label_colname = args$label_colname,
-        minimum_count_value_to_be_considered_nonzero = args$minimum_count_value_to_be_considered_nonzero,
-        minimum_number_of_samples_with_nonzero_counts_in_total = args$minimum_number_of_samples_with_nonzero_counts_in_total,
-        minimum_number_of_samples_with_nonzero_counts_in_a_group = args$minimum_number_of_samples_with_nonzero_counts_in_a_group,
-        use_cpm_counts_to_filter = args$use_cpm_counts_to_filter,
-        use_group_based_filtering = args$use_group_based_filtering,
-        plot_corr_matrix_heatmap = args$plot_corr_matrix_heatmap,
-        interactive_plots = args$interactive_plots
-        ) |> 
-    write_rds(file.path(getOption("moo_plots_dir"), "..", "moo", "moo.rds"))
+moo |>
+  filter_counts(
+    count_type = args$count_type,
+    feature_id_colname = args$feature_id_colname,
+    sample_id_colname = args$sample_id_colname,
+    group_colname = args$group_colname,
+    label_colname = args$label_colname,
+    minimum_count_value_to_be_considered_nonzero = args$minimum_count_value_to_be_considered_nonzero,
+    minimum_number_of_samples_with_nonzero_counts_in_total = args$minimum_number_of_samples_with_nonzero_counts_in_total,
+    minimum_number_of_samples_with_nonzero_counts_in_a_group = args$minimum_number_of_samples_with_nonzero_counts_in_a_group,
+    use_cpm_counts_to_filter = args$use_cpm_counts_to_filter,
+    use_group_based_filtering = args$use_group_based_filtering,
+    plot_corr_matrix_heatmap = args$plot_corr_matrix_heatmap,
+    interactive_plots = args$interactive_plots
+  ) |>
+  write_rds(file.path(getOption("moo_plots_dir"), "..", "moo", "moo-filt.rds"))

tests/test-setup.R

@@ -0,0 +1,45 @@
+#!/usr/bin/env Rscript
+
+# Simple test to verify the test setup and data exists
+cat("Testing basic test setup...\n")
+
+# Get the current working directory during test execution
+cwd <- getwd()
+cat("Current directory:", cwd, "\n")
+
+# Calculate repo root the same way the test does in tests/testthat/test-main.R
+# When running from tests/testthat, dirname(getwd()) goes to tests/
+repo_root_test_calculation <- normalizePath(file.path(dirname(cwd), ".."))
+cat(
+  "Calculated repo_root (test calculation):",
+  repo_root_test_calculation,
+  "\n"
+)
+
+# From script running in tests directory:
+# When this script is in tests/ directory, repo_root should be dirname(getwd())
+repo_root_script <- normalizePath(dirname(cwd))
+cat("Calculated repo_root (script location):", repo_root_script, "\n")
+
+# Check if test data exists using test calculation
+test_data_file <- file.path(repo_root_script, "data", "moo-batch.rds")
+cat("\nLooking for test data at:", test_data_file, "\n")
+cat("Test data exists:", file.exists(test_data_file), "\n")
+
+# Check code files
+code_main <- file.path(repo_root_script, "..", "code", "main.R")
+code_run <- file.path(repo_root_script, "..", "code", "run")
+cat("code/main.R exists:", file.exists(code_main), "\n")
+cat("code/run exists:", file.exists(code_run), "\n")
+
+# Try to load the MOO object
+if (file.exists(test_data_file)) {
+  cat("\nTrying to load MOO object...\n")
+  library(readr)
+  moo <- readr::read_rds(test_data_file)
+  cat("MOO object loaded successfully!\n")
+  cat("MOO class:", class(moo), "\n")
+  if (hasSlot(moo, "counts")) {
+    cat("MOO columns in counts:", names(moo@counts), "\n")
+  }
+}

tests/testthat.R

@@ -0,0 +1,2 @@
+library(testthat)
+test_dir(file.path('tests', 'testthat'), stop_on_failure = TRUE)

tests/testthat/test-main.R

@@ -0,0 +1,136 @@
+test_that("code/run executes successfully with default CLI arguments", {
+  # Create temporary workspace
+  workspace <- tempfile("mosuite_filter_counts_test_")
+  dir.create(workspace)
+  on.exit(unlink(workspace, recursive = TRUE), add = TRUE)
+
+  # Set up directory structure
+  code_dir <- file.path(workspace, "code")
+  data_dir <- file.path(workspace, "data")
+  results_dir <- file.path(code_dir, "..", "results")
+  dir.create(code_dir)
+  dir.create(data_dir)
+  dir.create(results_dir)
+
+  # Get test data from package tests directory
+  repo_root <- normalizePath(file.path(dirname(getwd()), ".."))
+  test_data_file <- file.path(repo_root, "tests", "data", "moo-clean.rds")
+
+  expect_true(
+    file.exists(test_data_file),
+    info = paste("Test data file should exist at", test_data_file)
+  )
+  file.copy(test_data_file, file.path(data_dir, "moo.rds"))
+
+  # Copy main.R and run script to workspace
+  file.copy(
+    file.path(repo_root, "code", "main.R"),
+    file.path(code_dir, "main.R")
+  )
+  file.copy(
+    file.path(repo_root, "code", "run"),
+    file.path(code_dir, "run")
+  )
+
+  # Run the script from code directory
+  old_wd <- getwd()
+  setwd(code_dir)
+  on.exit(setwd(old_wd), add = TRUE)
+
+  # Execute run script with default CLI arguments
+  exit_code <- system2(
+    "bash",
+    args = c(
+      "run",
+      "--count_type=clean",
+      "--minimum_count_value_to_be_considered_nonzero=8",
+      "--minimum_number_of_samples_with_nonzero_counts_in_total=7",
+      "--minimum_number_of_samples_with_nonzero_counts_in_a_group=3",
+      "--use_cpm_counts_to_filter=TRUE",
+      "--use_group_based_filtering=FALSE",
+      "--plot_corr_matrix_heatmap=FALSE"
+    )
+  )
+
+  # Check for successful execution
+  expect_equal(exit_code, 0, info = "run script should execute without error")
+  expect_true(
+    file.exists(file.path(results_dir, "moo", "moo-filt.rds")),
+    info = "Output file moo-filt.rds should be created"
+  )
+
+  # Validate output is a valid MOO object
+  moo <- readr::read_rds(file.path(results_dir, "moo", "moo-filt.rds"))
+  expect_true(
+    S7::S7_inherits(moo, MOSuite::multiOmicDataSet),
+    info = "Output should be an S7 multiOmicDataSet object"
+  )
+})
+
+test_that("code/run executes with custom CLI arguments", {
+  # Create temporary workspace
+  workspace <- tempfile("mosuite_filter_counts_custom_test_")
+  dir.create(workspace)
+  on.exit(unlink(workspace, recursive = TRUE), add = TRUE)
+
+  # Set up directory structure
+  code_dir <- file.path(workspace, "code")
+  data_dir <- file.path(workspace, "data")
+  results_dir <- file.path(code_dir, "..", "results")
+  dir.create(code_dir)
+  dir.create(data_dir)
+  dir.create(results_dir)
+
+  # Get test data from package tests directory
+  repo_root <- normalizePath(file.path(dirname(getwd()), ".."))
+  test_data_file <- file.path(repo_root, "tests", "data", "moo-clean.rds")
+
+  # Copy test data to workspace
+  file.copy(test_data_file, file.path(data_dir, "moo.rds"))
+
+  # Copy main.R and run script to workspace
+  file.copy(
+    file.path(repo_root, "code", "main.R"),
+    file.path(code_dir, "main.R")
+  )
+  file.copy(
+    file.path(repo_root, "code", "run"),
+    file.path(code_dir, "run")
+  )
+
+  # Run the script from code directory
+  old_wd <- getwd()
+  setwd(code_dir)
+  on.exit(setwd(old_wd), add = TRUE)
+
+  # Execute run script with custom CLI arguments
+  exit_code <- system2(
+    "bash",
+    args = c(
+      "run",
+      "--count_type=clean",
+      "--minimum_count_value_to_be_considered_nonzero=5",
+      "--minimum_number_of_samples_with_nonzero_counts_in_total=3",
+      "--use_cpm_counts_to_filter=FALSE",
+      "--plot_corr_matrix_heatmap=FALSE"
+    )
+  )
+
+  # Check for successful execution
+  expect_equal(
+    exit_code,
+    0,
+    info = "run script with custom args should execute without error"
+  )
+  expect_true(
+    file.exists(file.path(results_dir, "moo", "moo-filt.rds")),
+    info = "Output file moo-filt.rds should be created with custom args"
+  )
+
+  # Validate output is a valid MOO object
+  moo <- readr::read_rds(file.path(results_dir, "moo", "moo-filt.rds"))
+  expect_true(
+    S7::S7_inherits(moo, MOSuite::multiOmicDataSet),
+    info = "Output should be an S7 multiOmicDataSet object"
+  )
+})