MAINT Parameters validation for sklearn.datasets.fetch_rcv1 (#26126)

Charlie-XIAO · web-flow · commit c8fb5611823a · 2023-04-12T12:18:26.000+02:00
diff --git a/sklearn/datasets/_rcv1.py b/sklearn/datasets/_rcv1.py
@@ -26,6 +26,7 @@
 from ._svmlight_format_io import load_svmlight_files
 from ..utils import shuffle as shuffle_
 from ..utils import Bunch
+from ..utils._param_validation import validate_params, StrOptions
 
 
 # The original vectorized data can be found at:
@@ -76,6 +77,16 @@
 logger = logging.getLogger(__name__)
 
 
+@validate_params(
+    {
+        "data_home": [str, None],
+        "subset": [StrOptions({"train", "test", "all"})],
+        "download_if_missing": ["boolean"],
+        "random_state": ["random_state"],
+        "shuffle": ["boolean"],
+        "return_X_y": ["boolean"],
+    }
+)
 def fetch_rcv1(
     *,
     data_home=None,
diff --git a/sklearn/tests/test_public_functions.py b/sklearn/tests/test_public_functions.py
@@ -127,6 +127,7 @@ def _check_function_param_validation(
     "sklearn.datasets.fetch_lfw_pairs",
     "sklearn.datasets.fetch_lfw_people",
     "sklearn.datasets.fetch_olivetti_faces",
+    "sklearn.datasets.fetch_rcv1",
     "sklearn.datasets.load_svmlight_file",
     "sklearn.datasets.load_svmlight_files",
     "sklearn.datasets.make_biclusters",
