MAINT parameter validation for sklearn.datasets.load_svmlight_files (#25811)

genvalen · jeremiedbb · web-flow · commit 18aab046d383 · 2023-03-13T11:36:04.000+01:00
Co-authored-by: Jérémie du Boisberranger &lt;34657725+jeremiedbb@users.noreply.github.com&gt;
diff --git a/sklearn/datasets/_svmlight_format_io.py b/sklearn/datasets/_svmlight_format_io.py
@@ -245,6 +245,24 @@ def _open_and_load(f, dtype, multilabel, zero_based, query_id, offset=0, length=
     return data, indices, indptr, labels, query
 
 
+@validate_params(
+    {
+        "files": [
+            "array-like",
+            str,
+            os.PathLike,
+            HasMethods("read"),
+            Interval(Integral, 0, None, closed="left"),
+        ],
+        "n_features": [Interval(Integral, 1, None, closed="left"), None],
+        "dtype": "no_validation",  # delegate validation to numpy
+        "multilabel": ["boolean"],
+        "zero_based": ["boolean", StrOptions({"auto"})],
+        "query_id": ["boolean"],
+        "offset": [Interval(Integral, 0, None, closed="left")],
+        "length": [Integral],
+    }
+)
 def load_svmlight_files(
     files,
     *,
diff --git a/sklearn/tests/test_public_functions.py b/sklearn/tests/test_public_functions.py
@@ -109,6 +109,7 @@ def _check_function_param_validation(
     "sklearn.datasets.fetch_covtype",
     "sklearn.datasets.fetch_kddcup99",
     "sklearn.datasets.load_svmlight_file",
+    "sklearn.datasets.load_svmlight_files",
     "sklearn.datasets.make_classification",
     "sklearn.datasets.make_friedman1",
     "sklearn.datasets.make_sparse_coded_signal",
