@@ -1,6 +1,6 @@

 Package: fenr

 Title: Fast functional enrichment

-Version: 0.1.13

+Version: 0.1.14

 Authors@R: person(

     given = "Marek",

     family = "Gierlinski",

@@ -32,7 +32,6 @@ Imports:

     readr,

     stringr,

     tibble,

-    Rfast,

     httr,

     XML,

     jsonlite,

@@ -50,3 +50,6 @@

 - Added functions `get_term_features` and `get_feature_terms` to access data safely

 - HACK: BioPlanet server's SSL certificate expired, so need insecure download.

+## Version 0.1.14

+

+- Ditched large and clunky `Rfast` and using native R environments as fast hashes (see https://riptutorial.com/r/example/18339/environments-as-hash-maps)

@@ -77,10 +77,11 @@ prepare_for_enrichment <- function(terms, mapping, all_features = NULL, feature_

   }

   # Hash to select term name

-  term2name <- Rfast::Hash(

-    keys = terms$term_id,

-    values = terms$term_name

-  )

+  term2name <- new.env(hash = TRUE)

+  for (i in 1:nrow(terms)) {

+    r <- terms[i, ]

+    term2name[[r$term_id]] <- r$term_name

+  }

   # feature-term tibble

   feature_term <- mapping |>

@@ -93,18 +94,18 @@ prepare_for_enrichment <- function(terms, mapping, all_features = NULL, feature_

     dplyr::group_by(feature_id) |>

     dplyr::summarise(terms = list(term_id)) |>

     tibble::deframe()

-  feature2term <- Rfast::Hash()

+  feature2term <- new.env(hash = TRUE)

   for(feat in names(f2t))

-    feature2term[feat] <- f2t[[feat]]

+    feature2term[[feat]] <- f2t[[feat]]

   # Term to feature hash

   t2f <- feature_term |>

     dplyr::group_by(term_id) |>

     dplyr::summarise(features = list(feature_id)) |>

     tibble::deframe()

-  term2feature <- Rfast::Hash()

+  term2feature <- new.env(hash = TRUE)

   for(term in names(t2f))

-    term2feature[term] <- t2f[[term]]

+    term2feature[[term]] <- t2f[[term]]

   list(

     term2name = term2name,

@@ -171,7 +172,7 @@ functional_enrichment <- function(feat_all, feat_sel, term_data, feat2name = NUL

   # all terms present in the selection

   our_terms <- feat_sel |>

-    purrr::map(~term_data$feature2term[.x]) |>

+    purrr::map(~term_data$feature2term[[.x]]) |>

     unlist() |>

     unique()

@@ -182,8 +183,8 @@ functional_enrichment <- function(feat_all, feat_sel, term_data, feat2name = NUL

   res <- purrr::map_dfr(our_terms, function(term_id) {

     # all features with the term

-    # term_data$term2feature is a Hash object

-    tfeats <- term_data$term2feature[term_id]

+    # term_data$term2feature is a hash environment

+    tfeats <- term_data$term2feature[[term_id]]

     # features from selection with the term

     # this is faster than intersect(tfeats, feat_sel)

@@ -217,7 +218,7 @@ functional_enrichment <- function(feat_all, feat_sel, term_data, feat2name = NUL

     if (!is.null(feat2name)) tfeats_sel <- feat2name[tfeats_sel] |> unname()

-    term_name <- term_data$term2name[term_id]

+    term_name <- term_data$term2name[[term_id]]

     # returns NAs if no term found

     if (is.null(term_name)) term_name <- NA_character_

@@ -123,7 +123,7 @@ get_term_features <- function(term_data, term_id) {

   assert_that(is(term_data, "fenr_terms"))

   assert_that(is.string(term_id))

-  term_data$term2feature[term_id]

+  term_data$term2feature[[term_id]]

 }

@@ -142,5 +142,5 @@ get_feature_term <- function(term_data, feature_id) {

   assert_that(is(term_data, "fenr_terms"))

   assert_that(is.string(feature_id))

-  term_data$feature2term[feature_id]

+  term_data$feature2term[[feature_id]]

 }

@@ -1,5 +1,6 @@

 library(testthat)

+

 # Set 100 features

 N <- 100

 features_all <- sprintf("gene_%03d", seq_len(N))

@@ -33,7 +34,7 @@ test_that("Expected correct output", {

   # Check term names

   for(i in seq_along(terms$term_id)) {

     r <- terms[i, ]

-    expect_equal(sort(r$term_name), sort(td$term2name[r$term_id]))

+    expect_equal(sort(r$term_name), sort(td$term2name[[r$term_id]]))

   }

@@ -45,7 +46,7 @@ test_that("Expected correct output", {

         dplyr::filter(term_id == trm) |>

         dplyr::pull(feature_id) |>

         sort()

-      returned <- td$term2feature[trm] |>

+      returned <- td$term2feature[[trm]] |>

         sort()

       expect_equal(expected, returned)

     })

@@ -58,7 +59,7 @@ test_that("Expected correct output", {

         dplyr::filter(feature_id == feat) |>

         dplyr::pull(term_id) |>

         sort()

-      returned <- td$feature2term[feat] |>

+      returned <- td$feature2term[[feat]] |>

         sort()

       expect_equal(expected, returned)

     })