dataframe.astype() exception message to include column name (pandas-dev#48321)

* dataframe.astype() exception message to include column name

* added testcase

* review comments

* review comments

* review comments

* review comments

* Update doc/source/whatsnew/v1.6.0.rst

Co-authored-by: Matthew Roeschke <[email protected]>

* dataframe.astype() exception message to include column name

Co-authored-by: Matthew Roeschke <[email protected]>
Co-authored-by: Patrick Hoefler <[email protected]>

Author: 3 people

doc/source/whatsnew/v2.0.0.rst

@@ -36,7 +36,7 @@ Other enhancements
 - Added ``index`` parameter to :meth:`DataFrame.to_dict` (:issue:`46398`)
 - Added metadata propagation for binary operators on :class:`DataFrame` (:issue:`28283`)
 - :class:`.CategoricalConversionWarning`, :class:`.InvalidComparison`, :class:`.InvalidVersion`, :class:`.LossySetitemError`, and :class:`.NoBufferPresent` are now exposed in ``pandas.errors`` (:issue:`27656`)
--
+- :func:`DataFrame.astype` exception message thrown improved to include column name when type conversion is not possible. (:issue:`47571`)
 
 .. ---------------------------------------------------------------------------
 .. _whatsnew_200.notable_bug_fixes:

pandas/core/generic.py

@@ -6264,7 +6264,13 @@ def astype(
                 if isna(cdt):
                     res_col = col.copy() if copy else col
                 else:
-                    res_col = col.astype(dtype=cdt, copy=copy, errors=errors)
+                    try:
+                        res_col = col.astype(dtype=cdt, copy=copy, errors=errors)
+                    except ValueError as ex:
+                        ex.args = (
+                            f"{ex}: Error while type casting for column '{col_name}'",
+                        )
+                        raise
                 results.append(res_col)
 
         elif is_extension_array_dtype(dtype) and self.ndim > 1:

pandas/tests/frame/methods/test_astype.py

@@ -543,6 +543,18 @@ def test_astype_arg_for_errors(self):
 
         df.astype(np.int8, errors="ignore")
 
+    def test_astype_invalid_conversion(self):
+        # GH#47571
+        df = DataFrame({"a": [1, 2, "text"], "b": [1, 2, 3]})
+
+        msg = (
+            "invalid literal for int() with base 10: 'text': "
+            "Error while type casting for column 'a'"
+        )
+
+        with pytest.raises(ValueError, match=re.escape(msg)):
+            df.astype({"a": int})
+
     def test_astype_arg_for_errors_dictlist(self):
         # GH#25905
         df = DataFrame(