[R-package] allow use of categorical_features in Dataset when raw d…

…ata does not have column names (fixes #4374) (#5184)

* check for number of columns if data is matrixx for categorical indices check

* check for error when using a greater index than the number of columns

* apply suggestion

Co-authored-by: James Lamb <jaylamb20@gmail.com>

* revert whitespace change

* check if is filename instead of matrix

Co-authored-by: James Lamb <jaylamb20@gmail.com>

R-package/R/lgb.Dataset.R

@@ -169,12 +169,13 @@ Dataset <- R6::R6Class(
           } else {
 
             # Check if more categorical features were output over the feature space
-            if (max(private$categorical_feature) > length(private$colnames)) {
+            data_is_not_filename <- !is.character(private$raw_data)
+            if (data_is_not_filename && max(private$categorical_feature) > ncol(private$raw_data)) {
               stop(
                 "lgb.self.get.handle: supplied a too large value in categorical_feature: "
                 , max(private$categorical_feature)
                 , " but only "
-                , length(private$colnames)
+                , ncol(private$raw_data)
                 , " features"
               )
             }

R-package/tests/testthat/test_dataset.R

@@ -548,3 +548,18 @@ test_that("lgb.Dataset$get_feature_num_bin() works", {
   actual_num_bins <- sapply(1L:5L, ds$get_feature_num_bin)
   expect_identical(actual_num_bins, expected_num_bins)
 })
+
+test_that("lgb.Dataset can be constructed with categorical features and without colnames", {
+  # check that dataset can be constructed
+  raw_mat <- matrix(rep(c(0L, 1L), 50L), ncol = 1L)
+  ds <- lgb.Dataset(raw_mat, categorical_feature = 1L)$construct()
+  sparse_mat <- as(raw_mat, "dgCMatrix")
+  ds2 <- lgb.Dataset(sparse_mat, categorical_feature = 1L)$construct()
+  # check that the column names are NULL
+  expect_null(ds$.__enclos_env__$private$colnames)
+  expect_null(ds2$.__enclos_env__$private$colnames)
+  # check for error when index is greater than the number of columns
+  expect_error({
+    lgb.Dataset(raw_mat, categorical_feature = 2L)$construct()
+  }, regexp = "supplied a too large value in categorical_feature: 2 but only 1 features")
+})