feat: implement filter_low_counts()

resolves #2

.pre-commit-config.yaml

@@ -4,7 +4,8 @@ exclude: |
   (?x)(
       ^assets/|
       ^docs/.*.html|
-      ^data-raw/*.txt
+      ^data-raw/*.txt|
+      ^man/
    )
 repos:
   - repo: https://github.com/pre-commit/pre-commit-hooks

.prettierignore

@@ -7,3 +7,4 @@ results/
 *.code-workspace
 assets/*.html
 data-raw/*.txt
+man/*

DESCRIPTION

@@ -17,8 +17,9 @@ Depends:
     R (>= 2.10)
 Imports:
     dplyr,
-    readr
+    tidyr
 Suggests:
+    readr,
     testthat (>= 3.0.0)
 Config/testthat/edition: 3
 Encoding: UTF-8

NAMESPACE

@@ -1,4 +1,6 @@
 # Generated by roxygen2: do not edit by hand
 
+export("%>%")
 export(filter_low_counts)
 export(read_raw_counts)
+importFrom(dplyr,"%>%")

R/filter_low_counts.R

@@ -1,18 +1,27 @@
 #' filter_low_counts
 #'
-#' @param raw_counts_matrix raw_counts_matrix object
-#' @param min_counts integer number of min_counts across all samples, default 0
-#' @param min_cpm float minimum cpm value, default 0
-#' @param min_cpm_fraction float fraction of samples that need to satisfy min_cpm filter, default 1.0
+#' @param counts_dat dataframe of expected gene counts from RSEM
+#' @param min_counts integer number of minimum counts across all samples (default: 0)
 #'
-#' @return filtered_raw_count_matrix
+#' @return filtered counts dataframe
 #' @export
 #'
 #' @examples
+#' filter_low_counts(gene_counts)
 filter_low_counts <- function(
-    raw_counts_matrix,
-    min_counts = 0,
-    min_cpm = 0,
-    min_cpm_fraction = 1.0) {
-
+    counts_dat,
+    min_counts = 0) {
+  gene_id <- count <- count_sum <- NULL
+  genes_above_threshold <- counts_dat %>%
+    tidyr::pivot_longer(!c("gene_id", "GeneName"),
+      names_to = "sample_id", values_to = "count"
+    ) %>%
+    dplyr::group_by(gene_id) %>%
+    dplyr::summarize(count_sum = sum(count)) %>%
+    dplyr::filter(count_sum >= min_counts) %>%
+    dplyr::pull(gene_id)
+  return(
+    counts_dat %>%
+      dplyr::filter(gene_id %in% (genes_above_threshold))
+  )
 }

tests/testthat/test-filter_low_counts.R

@@ -1,3 +1,33 @@
-test_that("multiplication works", {
-  expect_equal(2 * 2, 4) # TODO write real tests
+test_that("filter_low_counts works", {
+  test_dat <- data.frame(
+    gene_id = c("A", "B", "C"),
+    GeneName = c("geneA", "geneB", "geneC"),
+    s1 = c(0, 0, 0),
+    s2 = c(0, 1, 0),
+    s3 = c(0, 0, 3)
+  )
+  expect_equal(filter_low_counts(test_dat), test_dat)
+  expect_equal(
+    filter_low_counts(test_dat, min_counts = 1),
+    data.frame(
+      gene_id = c("B", "C"),
+      GeneName = c("geneB", "geneC"),
+      s1 = c(0, 0), s2 = c(1, 0), s3 = c(0, 3)
+    )
+  )
+  expect_equal(
+    filter_low_counts(test_dat, min_counts = 2),
+    data.frame(
+      gene_id = "C", GeneName = "geneC",
+      s1 = 0, s2 = 0, s3 = 3
+    )
+  )
+  expect_equal(
+    filter_low_counts(test_dat, min_counts = 5),
+    data.frame(
+      gene_id = character(0),
+      GeneName = character(0),
+      s1 = numeric(0), s2 = numeric(0), s3 = numeric(0)
+    )
+  )
 })