Append and created a testthat unit test of the fix (#34)

* append and created a testthat unit test of the fix

* Append and created a testthat unit test of the fix

* made all the changes requested in the previous pull request

* fixed the documentation and deleted the r.data and r.history files

* updated the documentation and deleted rdata

* fixed the comments on documentation

* Remove documentation of assay as it is inherited by @inheritParam

* Fix unit tests that only occur on en_US.UTF-8

---------

Co-authored-by: Levi Waldron <[email protected]>

.gitignore

@@ -2,3 +2,5 @@
 lefser.Rproj
 docs
 vignettes/lefser_cache/
+.Rdata
+.Rhistory

DESCRIPTION

@@ -45,5 +45,5 @@ URL: https://github.com/waldronlab/lefser
 VignetteBuilder: knitr
 biocViews: Software, Sequencing, DifferentialExpression, Microbiome, 
   StatisticalMethod, Classification
-RoxygenNote: 7.2.3
+RoxygenNote: 7.3.1
 Roxygen: list(markdown = TRUE)

R/lefser.R

@@ -270,7 +270,7 @@ lefser <-
     } else {
         relab_data <- assay(relab, i = assay)
     }
-    if (checkAbundances && !identical(length(unique(colSums(relab_data))), 1L))
+    if (checkAbundances && !identical(all.equal(colSums(relab_data), rep(1e6, ncol(relab_data)), check.attributes = FALSE), TRUE))
         warning(
             "Convert counts to relative abundances with 'relativeAb()'"
         )

R/utils.R

@@ -20,16 +20,23 @@
 #' assay(se)
 #' assay(relativeAb(se))
 #'
+#' @description
+#' This function calculates the relative abundance of each feature in the SummarizedExperiment 
+#' object containing count data, expressed as counts per million (CPM) 
+#' 
+#' @returns returns a new SummarizedExperiment object with counts per million
+#' calculated and added as a new assay named rel_abs.
+#' 
 #' @export
 relativeAb <- function(se, assay = 1L) {
-    assay_data <- assay(se, i = assay)
-    csums <- colSums(assay_data)
-    div <- matrix(rep(csums, each = nrow(assay_data)), ncol = ncol(assay_data))
-    res <- assay_data / div
-    assaylist <- assays(se)
-    newalist <- append(
-        assaylist, values = S4Vectors::SimpleList(rel_abs = res), after = 0L
-    )
-    assays(se) <- newalist
-    se
+  assay_data <- assay(se, i = assay)
+  csums <- colSums(assay_data)
+  div <- matrix(rep(csums, each = nrow(assay_data)), ncol = ncol(assay_data))
+  res <- assay_data / div * 1e6
+  assaylist <- assays(se)
+  newalist <- append(
+    assaylist, values = S4Vectors::SimpleList(rel_abs = res), after = 0L
+  )
+  assays(se) <- newalist
+  se
 }

tests/testthat.R

@@ -1,4 +1,19 @@
 library(testthat)
 library(lefser)
 
+
+
 test_check("lefser")
+
+data(zeller14)
+
+zeller14sub <- zeller14[, zeller14$study_condition != "adenoma"]
+
+zeller14ra <- lefser::relativeAb(zeller14sub)
+
+
+expect_warning(lefser(zeller14sub, groupCol = "study_condition"), regexp = "^Convert counts.+")
+expect_equal(assayNames(zeller14ra), c("rel_abs", "exprs"))
+expect_equal(colSums(assay(zeller14ra, "rel_abs")), rep(1e6, ncol(zeller14ra)), check.attributes = FALSE)
+set.seed(1)
+expect_no_warning(x <- lefser(zeller14ra, groupCol = "study_condition"))

tests/testthat/test-lefser.R

@@ -11,7 +11,8 @@ test_that("lefser and lefserPlot work", {
   tol <- 0.2
   zellersub <- zeller14[1:150, zeller14$study_condition != "adenoma"]
   ## subsetting a DataFrame with NULL
-  expect_error(lefser(zellersub, groupCol = NULL, blockCol = NULL))
+  expect_warning(expect_error(lefser(zellersub, groupCol = NULL, blockCol = NULL)))
+  zellersub <- relativeAb(zellersub)
   results <- withr::with_seed(1,
     lefser(zellersub, groupCol = "study_condition", blockCol = NULL)
   )
@@ -21,8 +22,8 @@ test_that("lefser and lefserPlot work", {
     checkEnding(results, nrow(results), "Names", "o__Bacteroidales")
   )
 # TODO: compare results between LEfSe and lefser
-  expect_equal(results[1, "scores"], -6.431365, tolerance = tol)
-  expect_equal(results[nrow(results), "scores"], 6.402655, tolerance = tol)
+  expect_equal(results[1, "scores"], -3.688676, tolerance = tol)
+  expect_equal(results[nrow(results), "scores"], 3.815295, tolerance = tol)
   results2 <- withr::with_seed(1,
     lefser(
         zellersub, groupCol = "study_condition", blockCol = "age_category"
@@ -31,26 +32,29 @@ test_that("lefser and lefserPlot work", {
   expect_equal(colnames(results2), c("Names", "scores"))
   expect_true(checkEnding(results2, 1, "Names", "g__Ruminococcus"))
   expect_true(checkEnding(
-    results2, nrow(results2), "Names", "f__Clostridiales_noname"
+    results2, nrow(results2), "Names", "o__Bacteroidales"
   ))
-  expect_equal(results2[1, "scores"], -5.792313, tolerance = tol)
-  expect_equal(results2[nrow(results2), "scores"], 5.20564, tolerance = tol)
+  expect_equal(results2[1, "scores"], -3.58, tolerance = tol)
+  expect_equal(results2[nrow(results2), "scores"], 3.83, tolerance = tol)
 
-  expect_equal(nrow(results2), 6)
+  expect_equal(nrow(results2), 14)
   expect_true(all(
-    mapply(endsWith, results2$Names, c(
+    mapply(endsWith, results2$Names[1:7], c(
       "g__Ruminococcus",
       "o__Lactobacillales",
       "c__Bacilli",
+      "f__Streptococcaceae",
       "s__Ruminococcus_sp_5_1_39BFAA",
       "s__Eubacterium_hallii",
-      "f__Clostridiales_noname"
-    ))
+      "c__Deltaproteobacteria"
+      ))
   ))
 
   expect_equal(results2$scores,
-    c(-6.1384025872238, -5.90688120617861, -5.89755502841847,
-    -5.57735644006024, -5.42866570267924, 4.64159299932332),
+    c(-3.58274695616907, -3.32856514474989, -3.31328636163603, -3.28470366735756, 
+    -2.99971715869742, -2.82238742034892, 2.36885229314682, 2.36885229314682, 
+    2.36885229314682, 2.53194431507908, 2.53194431507908, 3.83089005744698, 
+    3.83093344936543, 3.83093344936543),
     tolerance = tol)
   plt <- lefserPlot(results2)
   expect_s3_class(plt, "ggplot")