scikit-learn · jeremiedbb · Oct 28, 2022 · Dec 28, 2020 · Dec 28, 2020 · Mar 10, 2022
diff --git a/doc/whats_new/v1.2.rst b/doc/whats_new/v1.2.rst
@@ -220,6 +220,11 @@ Changelog
 - |Enhancement| :func:`datasets.dump_svmlight_file` is now accelerated with a
   Cython implementation, providing 2-4x speedups.
   :pr:`23127` by :user:`Meekail Zain <micky774>`
+
+- |Enhancement| Path-like objects, such as those created with pathlib are now
+  allowed as paths in :func:`datasets.load_svmlight_file` and
+  :func:`datasets.load_svmlight_files`.
+  :pr:`19075` by :user:`Carlos Ramos Carreño <vnmabus>`
 
 :mod:`sklearn.decomposition`
 ............................

diff --git a/sklearn/datasets/_svmlight_format_io.py b/sklearn/datasets/_svmlight_format_io.py
@@ -85,12 +85,15 @@ def load_svmlight_file(
 
     Parameters
     ----------
-    f : str, file-like or int
+    f : str, path-like, file-like or int
         (Path to) a file to load. If a path ends in ".gz" or ".bz2", it will
         be uncompressed on the fly. If an integer is passed, it is assumed to
         be a file descriptor. A file-like or file descriptor will not be closed
         by this function. A file-like object must be opened in binary mode.
 
+        .. versionchanged:: 1.2
+           Path-like objects are now accepted.
+
     n_features : int, default=None
         The number of features to use. If None, it will be inferred. This
         argument is useful to load several files that are subsets of a
@@ -182,8 +185,10 @@ def get_data():
 def _gen_open(f):
     if isinstance(f, int):  # file descriptor
         return io.open(f, "rb", closefd=False)
+    elif isinstance(f, os.PathLike):
+        f = os.fspath(f)
     elif not isinstance(f, str):
-        raise TypeError("expected {str, int, file-like}, got %s" % type(f))
+        raise TypeError("expected {str, int, path-like, file-like}, got %s" % type(f))
 
     _, ext = os.path.splitext(f)
     if ext == ".gz":
@@ -249,13 +254,16 @@ def load_svmlight_files(
 
     Parameters
     ----------
-    files : array-like, dtype=str, file-like or int
+    files : array-like, dtype=str, path-like, file-like or int
         (Paths of) files to load. If a path ends in ".gz" or ".bz2", it will
         be uncompressed on the fly. If an integer is passed, it is assumed to
         be a file descriptor. File-likes and file descriptors will not be
         closed by this function. File-like objects must be opened in binary
         mode.
 
+        .. versionchanged:: 1.2
+           Path-like objects are now accepted.
+
     n_features : int, default=None
         The number of features to use. If None, it will be inferred from the
         maximum column index occurring in any of the files.

diff --git a/sklearn/datasets/tests/test_svmlight_format.py b/sklearn/datasets/tests/test_svmlight_format.py
@@ -11,7 +11,7 @@
 import pytest
 
 from sklearn.utils._testing import assert_array_equal
-from sklearn.utils._testing import assert_array_almost_equal
+from sklearn.utils._testing import assert_array_almost_equal, assert_allclose
 from sklearn.utils._testing import fails_if_pypy
 
 import sklearn
@@ -89,6 +89,16 @@ def test_load_svmlight_file_fd():
             os.close(fd)
 
 
+def test_load_svmlight_pathlib():
+    # test loading from file descriptor
+    with resources.path(TEST_DATA_MODULE, datafile) as data_path:
+        X1, y1 = load_svmlight_file(str(data_path))
+        X2, y2 = load_svmlight_file(data_path)
+
+    assert_allclose(X1.data, X2.data)
+    assert_allclose(y1, y2)
+
+
 def test_load_svmlight_file_multilabel():
     X, y = _load_svmlight_local_test_file(multifile, multilabel=True)
     assert y == [(0, 1), (2,), (), (1, 2)]