API Improves error msg when passing non-arrays to SimpleImputer (scikit-learn#14878)

Author: amueller

sklearn/impute/_base.py

@@ -182,9 +182,9 @@ def _validate_input(self, X):
                             force_all_finite=force_all_finite, copy=self.copy)
         except ValueError as ve:
             if "could not convert" in str(ve):
-                raise ValueError("Cannot use {0} strategy with non-numeric "
-                                 "data. Received datatype :{1}."
-                                 "".format(self.strategy, X.dtype.kind))
+                new_ve = ValueError("Cannot use {} strategy with non-numeric "
+                                    "data:\n{}".format(self.strategy, ve))
+                raise new_ve from None
             else:
                 raise ve
 

sklearn/impute/tests/test_impute.py

@@ -237,8 +237,23 @@ def test_imputation_mean_median_error_invalid_type(strategy, dtype):
     X = np.array([["a", "b", 3],
                   [4, "e", 6],
                   ["g", "h", 9]], dtype=dtype)
+    msg = "non-numeric data:\ncould not convert string to float: '"
+    with pytest.raises(ValueError, match=msg):
+        imputer = SimpleImputer(strategy=strategy)
+        imputer.fit_transform(X)
 
-    with pytest.raises(ValueError, match="non-numeric data"):
+
+@pytest.mark.parametrize("strategy", ["mean", "median"])
+@pytest.mark.parametrize("type", ['list', 'dataframe'])
+def test_imputation_mean_median_error_invalid_type_list_pandas(strategy, type):
+    X = [["a", "b", 3],
+         [4, "e", 6],
+         ["g", "h", 9]]
+    if type == 'dataframe':
+        pd = pytest.importorskip("pandas")
+        X = pd.DataFrame(X)
+    msg = "non-numeric data:\ncould not convert string to float: '"
+    with pytest.raises(ValueError, match=msg):
         imputer = SimpleImputer(strategy=strategy)
         imputer.fit_transform(X)