FIX Make GaussianNB more resilient

python/cuml/cuml/internals/array.py

@@ -985,7 +985,7 @@ def from_input(
             False (default) to not check at all.
 
         check_rows: boolean (default: False)
-            Set to an int `i` to check that input X has `i` columns. Set to
+            Set to an int `i` to check that input X has `i` rows. Set to
             False (default) to not check at all.
 
         fail_on_order: boolean (default: False)

python/cuml/cuml/internals/input_utils.py

@@ -333,7 +333,7 @@ def input_to_cuml_array(
         (default) to not check at all.
 
     check_rows: boolean (default: False)
-        Set to an int `i` to check that input X has `i` columns. Set to False
+        Set to an int `i` to check that input X has `i` rows. Set to False
         (default) to not check at all.
 
     fail_on_order: boolean (default: False)

python/cuml/cuml/naive_bayes/naive_bayes.py

@@ -368,6 +368,7 @@ def fit(self, X, y, sample_weight=None) -> "GaussianNB":
             Weights applied to individual samples (1. for unweighted).
             Currently sample weight is ignored.
         """
+        self.fit_called_ = False
         return self._partial_fit(
             X,
             y,
@@ -406,6 +407,7 @@ def _partial_fit(
         y = input_to_cupy_array(
             y,
             convert_to_dtype=(expected_y_dtype if convert_dtype else False),
+            check_rows=X.shape[0],
             check_dtype=expected_y_dtype,
         ).array
 

python/cuml/tests/test_sklearn_compatibility.py

@@ -674,6 +674,39 @@
         "check_fit1d": "HDBSCAN does not raise ValueError for 1D input",
         "check_fit2d_predict1d": "HDBSCAN does not handle 1D prediction input gracefully",
     },
+    GaussianNB: {
+        "check_estimator_tags_renamed": "No support for modern tags infrastructure",
+        "check_no_attributes_set_in_init": "GaussianNB sets attributes during init",
+        "check_dont_overwrite_parameters": "GaussianNB overwrites parameters during fit",
+        "check_estimators_unfitted": "GaussianNB does not raise NotFittedError before fit",
+        "check_do_not_raise_errors_in_init_or_set_params": "GaussianNB raises errors in init or set_params",
+        "check_n_features_in_after_fitting": "GaussianNB does not check n_features_in consistency",
+        "check_estimators_dtypes": "GaussianNB does not handle dtypes properly",
+        "check_sample_weights_pandas_series": "GaussianNB does not handle pandas Series sample weights",
+        "check_sample_weights_not_an_array": "GaussianNB does not handle non-array sample weights",
+        "check_sample_weights_shape": "GaussianNB does not validate sample weights shape",
+        "check_sample_weight_equivalence_on_dense_data": "GaussianNB sample weight equivalence not implemented",
+        "check_complex_data": "GaussianNB does not handle complex data",
+        "check_dtype_object": "GaussianNB does not handle object dtype",
+        "check_estimators_empty_data_messages": "GaussianNB does not handle empty data",
+        "check_estimators_nan_inf": "GaussianNB does not check for NaN and inf",
+        "check_estimator_sparse_tag": "GaussianNB does not support sparse data",
+        "check_estimator_sparse_array": "GaussianNB does not handle sparse arrays gracefully",
+        "check_classifier_data_not_an_array": "GaussianNB does not handle non-array data",
+        "check_classifiers_classes": "GaussianNB does not handle string data properly",
+        "check_estimators_partial_fit_n_features": "GaussianNB does not check n_features consistency in partial_fit",
+        "check_classifiers_train": "GaussianNB does not handle list inputs",
+        "check_classifiers_train(readonly_memmap=True)": "GaussianNB does not handle readonly memmap",
+        "check_classifiers_train(readonly_memmap=True,X_dtype=float32)": "GaussianNB does not handle readonly memmap with float32",
+        "check_classifiers_regression_target": "GaussianNB does not handle regression targets",
+        "check_supervised_y_no_nan": "GaussianNB does not check for NaN in y",
+        "check_supervised_y_2d": "GaussianNB does not handle 2D y",
+        "check_parameters_default_constructible": "GaussianNB parameters are mutated on init",
+        "check_fit_check_is_fitted": "GaussianNB passes check_is_fitted before being fit",
+        "check_fit1d": "GaussianNB does not raise ValueError for 1D input",
+        "check_fit2d_predict1d": "GaussianNB does not handle 1D prediction input gracefully",
+        "check_requires_y_none": "GaussianNB does not handle y=None",
+    },
 }
 
 
@@ -729,7 +762,7 @@ def test_sklearn_compatible_estimator(estimator, check):
     # estimators. As a result they are currently skipped.
     if isinstance(
         estimator,
-        (GaussianNB, ComplementNB, CategoricalNB, BernoulliNB, MultinomialNB),
+        (ComplementNB, CategoricalNB, BernoulliNB, MultinomialNB),
     ):
         pytest.skip(
             "Estimator leads to additional MemoryErrors in other estimators (gh-7100)"