[R-package] limit number of threads used in tests and examples (fixes #5987) (#5988)

R-package/man/lgb.plot.importance.Rd

@@ -47,6 +47,7 @@ params <- list(
     , learning_rate = 0.1
     , min_data_in_leaf = 1L
     , min_sum_hessian_in_leaf = 1.0
+    , num_threads = 2L
 )
 
 model <- lgb.train(

R-package/man/lgb.plot.interpretation.Rd

@@ -58,6 +58,7 @@ params <- list(
   , max_depth = -1L
   , min_data_in_leaf = 1L
   , min_sum_hessian_in_leaf = 1.0
+  , num_threads = 2L
 )
 model <- lgb.train(
   params = params

R-package/man/lgb.restore_handle.Rd

@@ -34,7 +34,9 @@ model <- lightgbm(
   , agaricus.train$label
   , params = list(objective = "binary")
   , nrounds = 5L
-  , verbose = 0)
+  , verbose = 0
+  , num_threads = 2L
+)
 fname <- tempfile(fileext="rds")
 saveRDS(model, fname)
 

R-package/man/lgb.save.Rd

@@ -33,6 +33,7 @@ params <- list(
   , metric = "l2"
   , min_data = 1L
   , learning_rate = 1.0
+  , num_threads = 2L
 )
 valids <- list(test = dtest)
 model <- lgb.train(

R-package/man/lgb.train.Rd

@@ -141,6 +141,7 @@ params <- list(
   , metric = "l2"
   , min_data = 1L
   , learning_rate = 1.0
+  , num_threads = 2L
 )
 valids <- list(test = dtest)
 model <- lgb.train(

R-package/man/predict.lgb.Booster.Rd

@@ -132,6 +132,7 @@ params <- list(
   , metric = "l2"
   , min_data = 1L
   , learning_rate = 1.0
+  , num_threads = 2L
 )
 valids <- list(test = dtest)
 model <- lgb.train(

R-package/man/readRDS.lgb.Booster.Rd

@@ -34,6 +34,7 @@ params <- list(
   , metric = "l2"
   , min_data = 1L
   , learning_rate = 1.0
+  , num_threads = 2L
 )
 valids <- list(test = dtest)
 model <- lgb.train(

R-package/man/saveRDS.lgb.Booster.Rd

@@ -57,6 +57,7 @@ params <- list(
   , metric = "l2"
   , min_data = 1L
   , learning_rate = 1.0
+  , num_threads = 2L
 )
 valids <- list(test = dtest)
 model <- lgb.train(

R-package/tests/testthat/helper.R

+# ref for this file:
+#
+# * https://r-pkgs.org/testing-design.html#testthat-setup-files
+
+# LightGBM-internal fix to comply with CRAN policy of only using up to 2 threads in tests and example.
+#
+# per https://cran.r-project.org/web/packages/policies.html
+#
+# > If running a package uses multiple threads/cores it must never use more than two simultaneously:
+#   the check farm is a shared resource and will typically be running many checks simultaneously.
+#
+.LGB_MAX_THREADS <- 2L

R-package/tests/testthat/test_Predictor.R

@@ -14,6 +14,7 @@ test_that("Predictor$finalize() should not fail", {
         data = dtrain
         , params = list(
             objective = "regression"
+            , num_threads = .LGB_MAX_THREADS
         )
         , verbose = VERBOSITY
         , nrounds = 3L
@@ -42,6 +43,7 @@ test_that("predictions do not fail for integer input", {
         data = dtrain
         , params = list(
             objective = "regression"
+            , num_threads = .LGB_MAX_THREADS
         )
         , verbose = VERBOSITY
         , nrounds = 3L
@@ -77,6 +79,7 @@ test_that("start_iteration works correctly", {
             , learning_rate = 0.6
             , objective = "binary"
             , verbosity = VERBOSITY
+            , num_threads = .LGB_MAX_THREADS
         )
         , nrounds = 50L
         , valids = list("test" = dtest)
@@ -126,7 +129,7 @@ test_that("Feature contributions from sparse inputs produce sparse outputs", {
       , obj = "regression"
       , nrounds = 5L
       , verbose = VERBOSITY
-      , params = list(min_data_in_leaf = 5L)
+      , params = list(min_data_in_leaf = 5L, num_threads = .LGB_MAX_THREADS)
     )
 
     pred_dense <- predict(bst, X, type = "contrib")
@@ -157,7 +160,7 @@ test_that("Sparse feature contribution predictions do not take inputs with wrong
       , obj = "regression"
       , nrounds = 5L
       , verbose = VERBOSITY
-      , params = list(min_data_in_leaf = 5L)
+      , params = list(min_data_in_leaf = 5L, num_threads = .LGB_MAX_THREADS)
     )
 
     X_wrong <- X[, c(1L:10L, 1L:10L)]
@@ -187,7 +190,7 @@ test_that("Feature contribution predictions do not take non-general CSR or CSC i
       , obj = "regression"
       , nrounds = 5L
       , verbose = VERBOSITY
-      , params = list(min_data_in_leaf = 5L)
+      , params = list(min_data_in_leaf = 5L, num_threads = .LGB_MAX_THREADS)
     )
 
     expect_error(predict(bst, SmatC, type = "contrib"))
@@ -211,6 +214,7 @@ test_that("predict() params should override keyword argument for raw-score predi
       objective = "binary"
       , min_data_in_leaf = 1L
       , seed = 708L
+      , num_threads = .LGB_MAX_THREADS
     )
     , nrounds = 10L
     , verbose = VERBOSITY
@@ -262,6 +266,7 @@ test_that("predict() params should override keyword argument for leaf-index pred
       objective = "regression"
       , min_data_in_leaf = 1L
       , seed = 708L
+      , num_threads = .LGB_MAX_THREADS
     )
     , nrounds = 10L
     , verbose = VERBOSITY
@@ -315,6 +320,7 @@ test_that("predict() params should override keyword argument for feature contrib
       objective = "regression"
       , min_data_in_leaf = 1L
       , seed = 708L
+      , num_threads = .LGB_MAX_THREADS
     )
     , nrounds = 10L
     , verbose = VERBOSITY
@@ -426,7 +432,7 @@ test_that("predict() keeps row names from data (regression)", {
         , obj = "regression"
         , nrounds = 5L
         , verbose = VERBOSITY
-        , params = list(min_data_in_leaf = 1L)
+        , params = list(min_data_in_leaf = 1L, num_threads = .LGB_MAX_THREADS)
     )
     .check_all_row_name_expectations(bst, X)
 })
@@ -442,6 +448,7 @@ test_that("predict() keeps row names from data (binary classification)", {
         , obj = "binary"
         , nrounds = 5L
         , verbose = VERBOSITY
+        , params = list(num_threads = .LGB_MAX_THREADS)
     )
     .check_all_row_name_expectations(bst, X)
 })
@@ -455,7 +462,7 @@ test_that("predict() keeps row names from data (multi-class classification)", {
     bst <- lgb.train(
         data = dtrain
         , obj = "multiclass"
-        , params = list(num_class = 3L)
+        , params = list(num_class = 3L, num_threads = .LGB_MAX_THREADS)
         , nrounds = 5L
         , verbose = VERBOSITY
     )
@@ -479,7 +486,7 @@ test_that("predictions for regression and binary classification are returned as
       , obj = "regression"
       , nrounds = 5L
       , verbose = VERBOSITY
-      , params = list(min_data_in_leaf = 1L)
+      , params = list(min_data_in_leaf = 1L, num_threads = .LGB_MAX_THREADS)
     )
     pred <- predict(model, X)
     expect_true(is.vector(pred))
@@ -497,6 +504,7 @@ test_that("predictions for regression and binary classification are returned as
       , obj = "binary"
       , nrounds = 5L
       , verbose = VERBOSITY
+      , params = list(num_threads = .LGB_MAX_THREADS)
     )
     pred <- predict(model, X)
     expect_true(is.vector(pred))
@@ -516,7 +524,7 @@ test_that("predictions for multiclass classification are returned as matrix", {
       , obj = "multiclass"
       , nrounds = 5L
       , verbose = VERBOSITY
-      , params = list(num_class = 3L)
+      , params = list(num_class = 3L, num_threads = .LGB_MAX_THREADS)
     )
     pred <- predict(model, X)
     expect_true(is.matrix(pred))
@@ -533,7 +541,7 @@ test_that("Single-row predictions are identical to multi-row ones", {
     X <- as.matrix(mtcars[, -1L])
     y <- mtcars[, 1L]
     dtrain <- lgb.Dataset(X, label = y, params = list(max_bin = 5L))
-    params <- list(min_data_in_leaf = 2L)
+    params <- list(min_data_in_leaf = 2L, num_threads = .LGB_MAX_THREADS)
     model <- lgb.train(
       params = params
      , data = dtrain
@@ -594,7 +602,7 @@ test_that("Fast-predict configuration accepts non-default prediction types", {
     X <- as.matrix(mtcars[, -1L])
     y <- mtcars[, 1L]
     dtrain <- lgb.Dataset(X, label = y, params = list(max_bin = 5L))
-    params <- list(min_data_in_leaf = 2L)
+    params <- list(min_data_in_leaf = 2L, num_threads = .LGB_MAX_THREADS)
     model <- lgb.train(
       params = params
      , data = dtrain
@@ -624,7 +632,7 @@ test_that("Fast-predict configuration does not block other prediction types", {
     X <- as.matrix(mtcars[, -1L])
     y <- mtcars[, 1L]
     dtrain <- lgb.Dataset(X, label = y, params = list(max_bin = 5L))
-    params <- list(min_data_in_leaf = 2L)
+    params <- list(min_data_in_leaf = 2L, num_threads = .LGB_MAX_THREADS)
     model <- lgb.train(
       params = params
      , data = dtrain
@@ -661,6 +669,7 @@ test_that("predict type='class' returns predicted class for classification objec
         , obj = "binary"
         , nrounds = 5L
         , verbose = VERBOSITY
+        , params = list(num_threads = .LGB_MAX_THREADS)
     )
     pred <- predict(bst, X, type = "class")
     expect_true(all(pred %in% c(0L, 1L)))
@@ -674,7 +683,7 @@ test_that("predict type='class' returns predicted class for classification objec
       , obj = "multiclass"
       , nrounds = 5L
       , verbose = VERBOSITY
-      , params = list(num_class = 3L)
+      , params = list(num_class = 3L, num_threads = .LGB_MAX_THREADS)
     )
     pred <- predict(model, X, type = "class")
     expect_true(all(pred %in% c(0L, 1L, 2L)))
@@ -690,6 +699,7 @@ test_that("predict type='class' returns values in the target's range for regress
         , obj = "regression"
         , nrounds = 5L
         , verbose = VERBOSITY
+        , params = list(num_threads = .LGB_MAX_THREADS)
     )
     pred <- predict(bst, X, type = "class")
     expect_true(!any(pred %in% c(0.0, 1.0)))

R-package/tests/testthat/test_custom_objective.R

@@ -39,6 +39,7 @@ param <- list(
   , objective = logregobj
   , metric = "auc"
   , verbose = VERBOSITY
+  , num_threads = .LGB_MAX_THREADS
 )
 num_round <- 10L
 
@@ -54,6 +55,7 @@ test_that("using a custom objective, custom eval, and no other metrics works", {
       num_leaves = 8L
       , learning_rate = 1.0
       , verbose = VERBOSITY
+      , num_threads = .LGB_MAX_THREADS
     )
     , data = dtrain
     , nrounds = 4L

R-package/tests/testthat/test_dataset.R

@@ -133,7 +133,7 @@ test_that("Dataset$set_reference() updates categorical_feature, colnames, and pr
   dtrain$construct()
   bst <- Booster$new(
     train_set = dtrain
-    , params = list(verbose = -1L)
+    , params = list(verbose = -1L, num_threads = .LGB_MAX_THREADS)
   )
   dtrain$.__enclos_env__$private$predictor <- bst$to_predictor()
 
@@ -394,6 +394,7 @@ test_that("lgb.Dataset: should be able to run lgb.train() immediately after usin
     , num_leaves = 5L
     , learning_rate = 1.0
     , verbose = VERBOSITY
+    , num_threads = .LGB_MAX_THREADS
   )
 
   # should be able to train right away
@@ -429,6 +430,7 @@ test_that("lgb.Dataset: should be able to run lgb.cv() immediately after using l
     , learning_rate = 1.0
     , num_iterations = 5L
     , verbosity = VERBOSITY
+    , num_threads = .LGB_MAX_THREADS
   )
 
   # should be able to train right away

R-package/tests/testthat/test_learning_to_rank.R

@@ -27,6 +27,7 @@ test_that("learning-to-rank with lgb.train() works as expected", {
         , lambdarank_truncation_level = 3L
         , learning_rate = 0.001
         , verbose = VERBOSITY
+        , num_threads = .LGB_MAX_THREADS
     )
     model <- lgb.train(
         params = params
@@ -91,6 +92,7 @@ test_that("learning-to-rank with lgb.cv() works as expected", {
         , min_data = 1L
         , learning_rate = 0.01
         , verbose = VERBOSITY
+        , num_threads = .LGB_MAX_THREADS
     )
     nfold <- 4L
     nrounds <- 10L

R-package/tests/testthat/test_lgb.Booster.R

@@ -13,6 +13,7 @@ test_that("Booster$finalize() should not fail", {
         data = dtrain
         , params = list(
             objective = "regression"
+            , num_threads = .LGB_MAX_THREADS
         )
         , verbose = VERBOSITY
         , nrounds = 3L
@@ -66,6 +67,7 @@ test_that("lgb.get.eval.result() should throw an informative error for incorrect
             , min_data = 1L
             , learning_rate = 1.0
             , verbose = VERBOSITY
+            , num_threads = .LGB_MAX_THREADS
         )
         , data = dtrain
         , nrounds = 5L
@@ -185,6 +187,7 @@ test_that("Loading a Booster from a text file works", {
         , learning_rate = 1.0
         , objective = "binary"
         , verbosity = VERBOSITY
+        , num_threads = .LGB_MAX_THREADS
     )
     bst <- lightgbm(
         data = as.matrix(train$data)
@@ -227,6 +230,7 @@ test_that("boosters with linear models at leaves can be written to text file and
         , metric = "mse"
         , seed = 0L
         , num_leaves = 2L
+        , num_threads = .LGB_MAX_THREADS
     )
 
     bst <- lgb.train(
@@ -268,6 +272,7 @@ test_that("Loading a Booster from a string works", {
             , learning_rate = 1.0
             , objective = "binary"
             , verbose = VERBOSITY
+            , num_threads = .LGB_MAX_THREADS
         )
         , nrounds = 2L
     )
@@ -299,6 +304,7 @@ test_that("Saving a large model to string should work", {
             num_leaves = 100L
             , learning_rate = 0.01
             , objective = "binary"
+            , num_threads = .LGB_MAX_THREADS
         )
         , nrounds = 500L
         , verbose = VERBOSITY
@@ -342,6 +348,7 @@ test_that("Saving a large model to JSON should work", {
             num_leaves = 100L
             , learning_rate = 0.01
             , objective = "binary"
+            , num_threads = .LGB_MAX_THREADS
         )
         , nrounds = 200L
         , verbose = VERBOSITY
@@ -372,6 +379,7 @@ test_that("If a string and a file are both passed to lgb.load() the file is used
             , learning_rate = 1.0
             , objective = "binary"
             , verbose = VERBOSITY
+            , num_threads = .LGB_MAX_THREADS
         )
         , nrounds = 2L
     )
@@ -406,6 +414,7 @@ test_that("Creating a Booster from a Dataset should work", {
         params = list(
             objective = "binary"
             , verbose = VERBOSITY
+            , num_threads = .LGB_MAX_THREADS
         ),
         train_set = dtrain
     )
@@ -427,6 +436,7 @@ test_that("Creating a Booster from a Dataset with an existing predictor should w
             , learning_rate = 1.0
             , objective = "binary"
             , verbose = VERBOSITY
+            , num_threads = .LGB_MAX_THREADS
         )
         , nrounds = nrounds
     )
@@ -440,6 +450,7 @@ test_that("Creating a Booster from a Dataset with an existing predictor should w
         train_set = dtest
         , params = list(
             verbose = VERBOSITY
+            , num_threads = .LGB_MAX_THREADS
         )
     )
     expect_true(lgb.is.Booster(bst))
@@ -463,6 +474,7 @@ test_that("Booster$eval() should work on a Dataset stored in a binary file", {
             , metric = "l2"
             , num_leaves = 4L
             , verbose = VERBOSITY
+            , num_threads = .LGB_MAX_THREADS
         )
         , data = dtrain
         , nrounds = 2L
@@ -492,7 +504,7 @@ test_that("Booster$eval() should work on a Dataset stored in a binary file", {
     eval_from_file <- bst$eval(
         data = lgb.Dataset(
             data = test_file
-            , params = list(verbose = VERBOSITY)
+            , params = list(verbose = VERBOSITY, num_threads = .LGB_MAX_THREADS)
         )$construct()
         , name = "test"
     )
@@ -521,6 +533,7 @@ test_that("Booster$rollback_one_iter() should work as expected", {
             , learning_rate = 1.0
             , objective = "binary"
             , verbose = VERBOSITY
+            , num_threads = .LGB_MAX_THREADS
         )
         , nrounds = nrounds
     )
@@ -555,6 +568,7 @@ test_that("Booster$update() passing a train_set works as expected", {
             , learning_rate = 1.0
             , objective = "binary"
             , verbose = VERBOSITY
+            , num_threads = .LGB_MAX_THREADS
         )
         , nrounds = nrounds
     )
@@ -579,6 +593,7 @@ test_that("Booster$update() passing a train_set works as expected", {
             , learning_rate = 1.0
             , objective = "binary"
             , verbose = VERBOSITY
+            , num_threads = .LGB_MAX_THREADS
         )
         , nrounds = nrounds +  1L
     )
@@ -604,6 +619,7 @@ test_that("Booster$update() throws an informative error if you provide a non-Dat
             , learning_rate = 1.0
             , objective = "binary"
             , verbose = VERBOSITY
+            , num_threads = .LGB_MAX_THREADS
         )
         , nrounds = nrounds
     )
@@ -631,6 +647,7 @@ test_that("Booster should store parameters and Booster$reset_parameter() should
         , boosting = "gbdt"
         , num_class = 5L
         , verbose = VERBOSITY
+        , num_threads = .LGB_MAX_THREADS
     )
     bst <- Booster$new(
         params = params
@@ -658,6 +675,7 @@ test_that("Booster$params should include dataset params, before and after Booste
         , max_depth = 4L
         , bagging_fraction = 0.8
         , verbose = VERBOSITY
+        , num_threads = .LGB_MAX_THREADS
     )
     bst <- Booster$new(
         params = params
@@ -670,6 +688,7 @@ test_that("Booster$params should include dataset params, before and after Booste
             , max_depth = 4L
             , bagging_fraction = 0.8
             , verbose = VERBOSITY
+            , num_threads = .LGB_MAX_THREADS
             , max_bin = 17L
         )
     )
@@ -681,6 +700,7 @@ test_that("Booster$params should include dataset params, before and after Booste
         , max_depth = 4L
         , bagging_fraction = 0.9
         , verbose = VERBOSITY
+        , num_threads = .LGB_MAX_THREADS
         , max_bin = 17L
     )
     expect_identical(ret_bst$params, expected_params)
@@ -699,6 +719,7 @@ test_that("Saving a model with different feature importance types works", {
             , learning_rate = 1.0
             , objective = "binary"
             , verbose = VERBOSITY
+            , num_threads = .LGB_MAX_THREADS
         )
         , nrounds = 2L
     )
@@ -754,6 +775,7 @@ test_that("Saving a model with unknown importance type fails", {
             , learning_rate = 1.0
             , objective = "binary"
             , verbose = VERBOSITY
+            , num_threads = .LGB_MAX_THREADS
         )
         , nrounds = 2L
     )
@@ -789,6 +811,7 @@ test_that("all parameters are stored correctly with save_model_to_string()", {
         params = list(
             objective = "regression"
             , metric = "l2"
+            , num_threads = .LGB_MAX_THREADS
         )
         , data = dtrain
         , nrounds = nrounds
@@ -845,6 +868,7 @@ test_that("early_stopping, num_iterations are stored correctly in model string e
         , n_iter = n_iter
         , early_stopping_round = early_stopping_round
         , n_iter_no_change = n_iter_no_change
+        , num_threads = .LGB_MAX_THREADS
     )
 
     bst <- lgb.train(
@@ -884,6 +908,7 @@ test_that("Booster: method calls Booster with a null handle should raise an info
             objective = "regression"
             , metric = "l2"
             , num_leaves = 8L
+            , num_threads = .LGB_MAX_THREADS
         )
         , data = dtrain
         , verbose = VERBOSITY
@@ -1074,6 +1099,7 @@ test_that("lgb.cv() correctly handles passing through params to the model file",
         , early_stopping_round = early_stopping_round
         , n_iter_no_change = n_iter_no_change
         , verbose = VERBOSITY
+        , num_threads = .LGB_MAX_THREADS
     )
 
     cv_bst <- lgb.cv(
@@ -1118,6 +1144,7 @@ test_that("params (including dataset params) should be stored in .rds file for B
         , max_depth = 4L
         , bagging_fraction = 0.8
         , verbose = VERBOSITY
+        , num_threads = .LGB_MAX_THREADS
     )
     bst <- Booster$new(
         params = params
@@ -1134,6 +1161,7 @@ test_that("params (including dataset params) should be stored in .rds file for B
             , max_depth = 4L
             , bagging_fraction = 0.8
             , verbose = VERBOSITY
+            , num_threads = .LGB_MAX_THREADS
             , max_bin = 17L
         )
     )
@@ -1153,6 +1181,7 @@ test_that("params (including dataset params) should be stored in .rds file for B
         , max_depth = 4L
         , bagging_fraction = 0.8
         , verbose = VERBOSITY
+        , num_threads = .LGB_MAX_THREADS
     )
     bst <- Booster$new(
         params = params
@@ -1169,6 +1198,7 @@ test_that("params (including dataset params) should be stored in .rds file for B
             , max_depth = 4L
             , bagging_fraction = 0.8
             , verbose = VERBOSITY
+            , num_threads = .LGB_MAX_THREADS
             , max_bin = 17L
         )
     )
@@ -1184,6 +1214,7 @@ test_that("Handle is automatically restored when calling predict", {
         , params = list(
             verbose = VERBOSITY
         )
+        , num_threads = .LGB_MAX_THREADS
     )
     bst_file <- tempfile(fileext = ".rds")
     saveRDS(bst, file = bst_file)
@@ -1209,6 +1240,7 @@ test_that("boosters with linear models at leaves work with saveRDS.lgb.Booster a
         , metric = "mse"
         , seed = 0L
         , num_leaves = 2L
+        , num_threads = .LGB_MAX_THREADS
     )
 
     bst <- lgb.train(
@@ -1248,6 +1280,7 @@ test_that("boosters with linear models at leaves can be written to RDS and re-lo
         , metric = "mse"
         , seed = 0L
         , num_leaves = 2L
+        , num_threads = .LGB_MAX_THREADS
     )
 
     bst <- lgb.train(
@@ -1344,6 +1377,7 @@ test_that("Booster's print, show, and summary work correctly", {
         params = list(
           objective = "regression"
           , min_data_in_leaf = 1L
+          , num_threads = .LGB_MAX_THREADS
         )
         , data = lgb.Dataset(
             as.matrix(mtcars[, -1L])
@@ -1359,7 +1393,7 @@ test_that("Booster's print, show, and summary work correctly", {
 
     data("iris")
     model <- lgb.train(
-        params = list(objective = "multiclass", num_class = 3L)
+        params = list(objective = "multiclass", num_class = 3L, num_threads = .LGB_MAX_THREADS)
         , data = lgb.Dataset(
             as.matrix(iris[, -5L])
             , label = as.numeric(factor(iris$Species)) - 1.0
@@ -1399,6 +1433,7 @@ test_that("Booster's print, show, and summary work correctly", {
         , eval = .evalerror
         , verbose = VERBOSITY
         , nrounds = 5L
+        , params = list(num_threads = .LGB_MAX_THREADS)
     )
 
     .check_methods_work(model)
@@ -1410,6 +1445,7 @@ test_that("LGBM_BoosterGetNumFeature_R returns correct outputs", {
         params = list(
           objective = "regression"
           , min_data_in_leaf = 1L
+          , num_threads = .LGB_MAX_THREADS
         )
         , data = lgb.Dataset(
             as.matrix(mtcars[, -1L])

R-package/tests/testthat/test_lgb.interprete.R

@@ -31,6 +31,7 @@ test_that("lgb.intereprete works as expected for binary classification", {
         , min_data_in_leaf = 1L
         , min_sum_hessian_in_leaf = 1.0
         , verbose = VERBOSITY
+        , num_threads = .LGB_MAX_THREADS
     )
     model <- lgb.train(
         params = params
@@ -83,6 +84,7 @@ test_that("lgb.intereprete works as expected for multiclass classification", {
         , learning_rate = 0.00001
         , min_data = 1L
         , verbose = VERBOSITY
+        , num_threads = .LGB_MAX_THREADS
     )
     model <- lgb.train(
         params = params

R-package/tests/testthat/test_lgb.plot.importance.R

@@ -14,6 +14,7 @@ test_that("lgb.plot.importance() should run without error for well-formed inputs
         , min_data_in_leaf = 1L
         , min_sum_hessian_in_leaf = 1.0
         , verbosity = VERBOSITY
+        , num_threads = .LGB_MAX_THREADS
     )
     model <- lgb.train(params, dtrain, 3L)
     tree_imp <- lgb.importance(model, percentage = TRUE)

R-package/tests/testthat/test_lgb.plot.interpretation.R

@@ -31,6 +31,7 @@ test_that("lgb.plot.interepretation works as expected for binary classification"
         , min_data_in_leaf = 1L
         , min_sum_hessian_in_leaf = 1.0
         , verbosity = VERBOSITY
+        , num_threads = .LGB_MAX_THREADS
     )
     model <- lgb.train(
         params = params
@@ -80,6 +81,7 @@ test_that("lgb.plot.interepretation works as expected for multiclass classificat
         , num_class = 3L
         , learning_rate = 0.00001
         , min_data = 1L
+        , num_threads = .LGB_MAX_THREADS
     )
     model <- lgb.train(
         params = params

R-package/tests/testthat/test_parameters.R

@@ -20,6 +20,7 @@ test_that("Feature penalties work properly", {
         , objective = "binary"
         , feature_penalty = paste0(feature_penalties, collapse = ",")
         , metric = "binary_error"
+        , num_threads = .LGB_MAX_THREADS
       )
       , nrounds = 5L
       , verbose = -1L
@@ -97,6 +98,7 @@ test_that("training should warn if you use 'dart' boosting, specified with 'boos
         , learning_rate = 0.05
         , objective = "binary"
         , metric = "binary_error"
+        , num_threads = .LGB_MAX_THREADS
     )
     params[[boosting_param]] <- "dart"
     expect_warning({

R-package/tests/testthat/test_weighted_loss.R

@@ -9,7 +9,7 @@ test_that("Gamma regression reacts on 'weight'", {
   y <- X[, 1L] + X[, 2L] + runif(n)
   X_pred <- X[1L:5L, ]
 
-  params <- list(objective = "gamma")
+  params <- list(objective = "gamma", num_threads = .LGB_MAX_THREADS)
 
   # Unweighted
   dtrain <- lgb.Dataset(X, label = y)