Change basic_statistics API for 2025.0 (#2026)

* Update basic_statistics.py

* Update test_basic_statistics.py

* Update basic_statistics.py

Author: icfaust

onedal/basic_statistics/basic_statistics.py

@@ -96,27 +96,6 @@ def fit(self, data, sample_weight=None, queue=None):
 
         return self
 
-    def compute(self, data, weights=None, queue=None):
-        warnings.warn(
-            "Method `compute` was deprecated in version 2024.7 and will be "
-            "removed in 2025.0. Use `fit` instead."
-        )
-
-        is_csr = _is_csr(data)
-
-        if data is not None:
-            data = _check_array(data, ensure_2d=False)
-        if weights is not None:
-            weights = _check_array(weights, ensure_2d=False)
-
-        policy = self._get_policy(queue, data, weights)
-        data, weights = _convert_to_supported(policy, data, weights)
-        data_table, weights_table = to_table(data, weights)
-        dtype = data.dtype
-        res = self._compute_raw(data_table, weights_table, policy, dtype, is_csr)
-
-        return {k: from_table(v).ravel() for k, v in res.items()}
-
     def _compute_raw(
         self, data_table, weights_table, policy, dtype=np.float32, is_csr=False
     ):

onedal/basic_statistics/tests/test_basic_statistics.py

@@ -296,19 +296,3 @@ def test_options_csr(queue, option, dtype):
     tol = fp32tol if res.dtype == np.float32 else fp64tol
 
     assert_allclose(gtr, res, rtol=tol)
-
-
-def test_warning():
-    basicstat = BasicStatistics()
-    data = np.array([0, 1])
-
-    with pytest.warns(
-        UserWarning,
-        match="Method `compute` was deprecated in version 2024.7 and will be removed in 2025.0. Use `fit` instead.",
-    ) as warn_record:
-        basicstat.compute(data)
-
-    if daal_check_version((2025, "P", 0)):
-        assert len(warn_record) == 0
-    else:
-        assert len(warn_record) == 1

sklearnex/basic_statistics/basic_statistics.py

@@ -107,9 +107,6 @@ def _onedal_fit(self, X, sample_weight=None, queue=None):
         self._onedal_estimator.fit(X, sample_weight, queue)
         self._save_attributes()
 
-    def compute(self, data, weights=None, queue=None):
-        return self._onedal_estimator.compute(data, weights, queue)
-
     def fit(self, X, y=None, *, sample_weight=None):
         """Compute statistics with X, using minibatches of size batch_size.