Standardize input validation error type on pandas

Author: Yikun

python/pyspark/pandas/base.py

@@ -1498,7 +1498,7 @@ def shift(self, periods=1, fill_value=None) -> Union["Series", "Index"]:
 
     def _shift(self, periods, fill_value, *, part_cols=()):
         if not isinstance(periods, int):
-            raise ValueError("periods should be an int; however, got [%s]" % type(periods).__name__)
+            raise TypeError("periods should be an int; however, got [%s]" % type(periods).__name__)
 
         col = self.spark.column
         window = (
@@ -1828,7 +1828,7 @@ def take(self, indices) -> Union["Series", "Index"]:
                    )
         """
         if not is_list_like(indices) or isinstance(indices, (dict, set)):
-            raise ValueError("`indices` must be a list-like except dict or set")
+            raise TypeError("`indices` must be a list-like except dict or set")
         if isinstance(self, ps.Series):
             return cast(ps.Series, self.iloc[indices])
         else:

python/pyspark/pandas/config.py

@@ -70,7 +70,7 @@ class Option:
     >>> option.validate('abc')  # doctest: +NORMALIZE_WHITESPACE
     Traceback (most recent call last):
       ...
-    ValueError: The value for option 'option.name' was <class 'str'>;
+    TypeError: The value for option 'option.name' was <class 'str'>;
     however, expected types are [(<class 'float'>, <class 'int'>)].
 
     >>> option.validate(-1.1)
@@ -101,7 +101,7 @@ def validate(self, v: Any) -> None:
         Validate the given value and throw an exception with related information such as key.
         """
         if not isinstance(v, self.types):
-            raise ValueError(
+            raise TypeError(
                 "The value for option '%s' was %s; however, expected types are "
                 "[%s]." % (self.key, type(v), str(self.types))
             )

python/pyspark/pandas/frame.py

@@ -771,7 +771,7 @@ def _map_series_op(self, op, other):
         if not isinstance(other, DataFrame) and (
             isinstance(other, IndexOpsMixin) or is_sequence(other)
         ):
-            raise ValueError(
+            raise TypeError(
                 "%s with a sequence is currently not supported; "
                 "however, got %s." % (op, type(other).__name__)
             )
@@ -2936,7 +2936,7 @@ class  locomotion
         from pyspark.pandas.series import first_series
 
         if not is_name_like_value(key):
-            raise ValueError("'key' should be a scalar value or tuple that contains scalar values")
+            raise TypeError("'key' should be a scalar value or tuple that contains scalar values")
 
         if level is not None and is_name_like_tuple(key):
             raise KeyError(key)
@@ -3301,7 +3301,7 @@ def where(self, cond, other=np.nan) -> "DataFrame":
             ]
             kdf[tmp_cond_col_names] = cond
         else:
-            raise ValueError("type of cond must be a DataFrame or Series")
+            raise TypeError("type of cond must be a DataFrame or Series")
 
         tmp_other_col_names = [
             tmp_other_col_name(name_like_string(label)) for label in self._internal.column_labels
@@ -3431,7 +3431,7 @@ def mask(self, cond, other=np.nan) -> "DataFrame":
         from pyspark.pandas.series import Series
 
         if not isinstance(cond, (DataFrame, Series)):
-            raise ValueError("type of cond must be a DataFrame or Series")
+            raise TypeError("type of cond must be a DataFrame or Series")
 
         cond_inversed = cond._apply_series_op(lambda kser: ~kser)
         return self.where(cond_inversed, other)
@@ -3997,7 +3997,7 @@ def insert(
         assert allow_duplicates is False
 
         if not is_name_like_value(column):
-            raise ValueError(
+            raise TypeError(
                 '"column" should be a scalar value or tuple that contains scalar values'
             )
 
@@ -4289,7 +4289,7 @@ def round(self, decimals=0) -> "DataFrame":
         elif isinstance(decimals, int):
             decimals = {k: decimals for k in self._internal.column_labels}
         else:
-            raise ValueError("decimals must be an integer, a dict-like or a Series")
+            raise TypeError("decimals must be an integer, a dict-like or a Series")
 
         def op(kser):
             label = kser._column_label
@@ -5660,7 +5660,7 @@ def clip(self, lower: Union[float, int] = None, upper: Union[float, int] = None)
         will output the original DataFrame, simply ignoring the incompatible types.
         """
         if is_list_like(lower) or is_list_like(upper):
-            raise ValueError(
+            raise TypeError(
                 "List-like value are not supported for 'lower' and 'upper' at the " + "moment"
             )
 
@@ -5941,20 +5941,20 @@ def pivot_table(
         small  5.5  2.333333  17  13
         """
         if not is_name_like_value(columns):
-            raise ValueError("columns should be one column name.")
+            raise TypeError("columns should be one column name.")
 
         if not is_name_like_value(values) and not (
             isinstance(values, list) and all(is_name_like_value(v) for v in values)
         ):
-            raise ValueError("values should be one column or list of columns.")
+            raise TypeError("values should be one column or list of columns.")
 
         if not isinstance(aggfunc, str) and (
             not isinstance(aggfunc, dict)
             or not all(
                 is_name_like_value(key) and isinstance(value, str) for key, value in aggfunc.items()
             )
         ):
-            raise ValueError(
+            raise TypeError(
                 "aggfunc must be a dict mapping from column name "
                 "to aggregate functions (string)."
             )
@@ -6031,7 +6031,7 @@ def pivot_table(
                 .agg(*agg_cols)
             )
         else:
-            raise ValueError("index should be a None or a list of columns.")
+            raise TypeError("index should be a None or a list of columns.")
 
         if fill_value is not None and isinstance(fill_value, (int, float)):
             sdf = sdf.fillna(fill_value)
@@ -7940,7 +7940,7 @@ def append(
         3  3  4
         """
         if isinstance(other, ps.Series):
-            raise ValueError("DataFrames.append() does not support appending Series to DataFrames")
+            raise TypeError("DataFrames.append() does not support appending Series to DataFrames")
         if sort:
             raise NotImplementedError("The 'sort' parameter is currently not supported")
 
@@ -10726,7 +10726,7 @@ def quantile(
             raise NotImplementedError('axis should be either 0 or "index" currently.')
 
         if not isinstance(accuracy, int):
-            raise ValueError(
+            raise TypeError(
                 "accuracy must be an integer; however, got [%s]" % type(accuracy).__name__
             )
 
@@ -10735,7 +10735,7 @@ def quantile(
 
         for v in q if isinstance(q, list) else [q]:
             if not isinstance(v, float):
-                raise ValueError(
+                raise TypeError(
                     "q must be a float or an array of floats; however, [%s] found." % type(v)
                 )
             if v < 0.0 or v > 1.0:
@@ -10904,9 +10904,9 @@ def query(self, expr, inplace=False) -> Optional["DataFrame"]:
         0  1  10   10
         """
         if isinstance(self.columns, pd.MultiIndex):
-            raise ValueError("Doesn't support for MultiIndex columns")
+            raise TypeError("Doesn't support for MultiIndex columns")
         if not isinstance(expr, str):
-            raise ValueError(
+            raise TypeError(
                 "expr must be a string to be evaluated, {} given".format(type(expr).__name__)
             )
         inplace = validate_bool_kwarg(inplace, "inplace")
@@ -11012,7 +11012,7 @@ class  max_speed
         """
         axis = validate_axis(axis)
         if not is_list_like(indices) or isinstance(indices, (dict, set)):
-            raise ValueError("`indices` must be a list-like except dict or set")
+            raise TypeError("`indices` must be a list-like except dict or set")
         if axis == 0:
             return cast(DataFrame, self.iloc[indices, :])
         else:
@@ -11098,7 +11098,7 @@ def eval(self, expr, inplace=False) -> Optional[Union["DataFrame", "Series"]]:
         from pyspark.pandas.series import first_series
 
         if isinstance(self.columns, pd.MultiIndex):
-            raise ValueError("`eval` is not supported for multi-index columns")
+            raise TypeError("`eval` is not supported for multi-index columns")
         inplace = validate_bool_kwarg(inplace, "inplace")
         should_return_series = False
         series_name = None
@@ -11179,7 +11179,7 @@ def explode(self, column) -> "DataFrame":
         from pyspark.pandas.series import Series
 
         if not is_name_like_value(column):
-            raise ValueError("column must be a scalar")
+            raise TypeError("column must be a scalar")
 
         kdf = DataFrame(self._internal.resolved_copy)  # type: "DataFrame"
         kser = kdf[column]

python/pyspark/pandas/generic.py

@@ -1895,7 +1895,7 @@ def median(
             numeric_only = True
 
         if not isinstance(accuracy, int):
-            raise ValueError(
+            raise TypeError(
                 "accuracy must be an integer; however, got [%s]" % type(accuracy).__name__
             )
 

python/pyspark/pandas/groupby.py

@@ -2416,7 +2416,7 @@ def median(self, numeric_only=True, accuracy=10000) -> Union[DataFrame, Series]:
         Name: b, dtype: float64
         """
         if not isinstance(accuracy, int):
-            raise ValueError(
+            raise TypeError(
                 "accuracy must be an integer; however, got [%s]" % type(accuracy).__name__
             )
 

python/pyspark/pandas/indexes/base.py

@@ -2076,7 +2076,7 @@ def repeat(self, repeats: int) -> "Index":
         MultiIndex([], )
         """
         if not isinstance(repeats, int):
-            raise ValueError(
+            raise TypeError(
                 "`repeats` argument must be integer, but got {}".format(type(repeats).__name__)
             )
         elif repeats < 0:

python/pyspark/pandas/indexes/multi.py

@@ -342,7 +342,7 @@ def from_frame(df, names=None) -> "MultiIndex":
         if names is None:
             names = df._internal.column_labels
         elif not is_list_like(names):
-            raise ValueError("Names should be list-like for a MultiIndex")
+            raise TypeError("Names should be list-like for a MultiIndex")
         else:
             names = [name if is_name_like_tuple(name) else (name,) for name in names]
 

python/pyspark/pandas/namespace.py

@@ -126,7 +126,7 @@ def from_pandas(pobj: Union[pd.DataFrame, pd.Series, pd.Index]) -> Union[Series,
     elif isinstance(pobj, pd.Index):
         return DataFrame(pd.DataFrame(index=pobj)).index
     else:
-        raise ValueError("Unknown data type: {}".format(type(pobj).__name__))
+        raise TypeError("Unknown data type: {}".format(type(pobj).__name__))
 
 
 _range = range  # built-in range
@@ -2770,7 +2770,7 @@ def broadcast(obj) -> DataFrame:
     ...
     """
     if not isinstance(obj, DataFrame):
-        raise ValueError("Invalid type : expected DataFrame got {}".format(type(obj).__name__))
+        raise TypeError("Invalid type : expected DataFrame got {}".format(type(obj).__name__))
     return DataFrame(
         obj._internal.with_new_sdf(F.broadcast(obj._internal.resolved_copy.spark_frame))
     )

python/pyspark/pandas/plot/core.py

@@ -40,7 +40,7 @@ def get_top_n(self, data):
         if isinstance(data, (Series, DataFrame)):
             data = data.head(max_rows + 1).to_pandas()
         else:
-            raise ValueError("Only DataFrame and Series are supported for plotting.")
+            raise TypeError("Only DataFrame and Series are supported for plotting.")
 
         self.partial = False
         if len(data) > max_rows:
@@ -80,7 +80,7 @@ def get_sampled(self, data):
             sampled = data._internal.resolved_copy.spark_frame.sample(fraction=self.fraction)
             return DataFrame(data._internal.with_new_sdf(sampled)).to_pandas()
         else:
-            raise ValueError("Only DataFrame and Series are supported for plotting.")
+            raise TypeError("Only DataFrame and Series are supported for plotting.")
 
     def set_result_text(self, ax):
         assert hasattr(self, "fraction")

python/pyspark/pandas/series.py

@@ -2016,7 +2016,7 @@ def clip(self, lower: Union[float, int] = None, upper: Union[float, int] = None)
         original Series, simply ignoring the incompatible types.
         """
         if is_list_like(lower) or is_list_like(upper):
-            raise ValueError(
+            raise TypeError(
                 "List-like value are not supported for 'lower' and 'upper' at the " + "moment"
             )
 
@@ -3182,7 +3182,7 @@ def aggregate(self, func: Union[str, List[str]]) -> Union[Scalar, "Series"]:
         elif isinstance(func, str):
             return getattr(self, func)()
         else:
-            raise ValueError("func must be a string or list of strings")
+            raise TypeError("func must be a string or list of strings")
 
     agg = aggregate
 
@@ -3345,7 +3345,7 @@ def round(self, decimals=0) -> "Series":
         Name: x, dtype: float64
         """
         if not isinstance(decimals, int):
-            raise ValueError("decimals must be an integer")
+            raise TypeError("decimals must be an integer")
         scol = F.round(self.spark.column, decimals)
         return self._with_new_scol(scol)
 
@@ -3402,12 +3402,12 @@ def quantile(
             ).rename(self.name)
         else:
             if not isinstance(accuracy, int):
-                raise ValueError(
+                raise TypeError(
                     "accuracy must be an integer; however, got [%s]" % type(accuracy).__name__
                 )
 
             if not isinstance(q, float):
-                raise ValueError(
+                raise TypeError(
                     "q must be a float or an array of floats; however, [%s] found." % type(q)
                 )
             if q < 0.0 or q > 1.0:
@@ -3639,7 +3639,7 @@ def diff(self, periods=1) -> "Series":
 
     def _diff(self, periods, *, part_cols=()):
         if not isinstance(periods, int):
-            raise ValueError("periods should be an int; however, got [%s]" % type(periods).__name__)
+            raise TypeError("periods should be an int; however, got [%s]" % type(periods).__name__)
         window = (
             Window.partitionBy(*part_cols)
             .orderBy(NATURAL_ORDER_COLUMN_NAME)
@@ -3984,7 +3984,7 @@ def pop(self, item) -> Union["Series", Scalar]:
         dtype: float64
         """
         if not is_name_like_value(item):
-            raise ValueError("'key' should be string or tuple that contains strings")
+            raise TypeError("'key' should be string or tuple that contains strings")
         if not is_name_like_tuple(item):
             item = (item,)
         if self._internal.index_level < len(item):
@@ -4328,7 +4328,7 @@ def replace(self, to_replace=None, value=None, regex=False) -> "Series":
         if to_replace is None:
             return self.fillna(method="ffill")
         if not isinstance(to_replace, (str, list, tuple, dict, int, float)):
-            raise ValueError("'to_replace' should be one of str, list, tuple, dict, int, float")
+            raise TypeError("'to_replace' should be one of str, list, tuple, dict, int, float")
         if regex:
             raise NotImplementedError("replace currently not support for regex")
         to_replace = list(to_replace) if isinstance(to_replace, tuple) else to_replace
@@ -4438,7 +4438,7 @@ def update(self, other) -> None:
         >>> reset_option("compute.ops_on_diff_frames")
         """
         if not isinstance(other, Series):
-            raise ValueError("'other' must be a Series")
+            raise TypeError("'other' must be a Series")
 
         combined = combine_frames(self._kdf, other._kdf, how="leftouter")
 
@@ -4813,7 +4813,7 @@ def combine_first(self, other) -> "Series":
         dtype: float64
         """
         if not isinstance(other, ps.Series):
-            raise ValueError("`combine_first` only allows `Series` for parameter `other`")
+            raise TypeError("`combine_first` only allows `Series` for parameter `other`")
         if same_anchor(self, other):
             this = self.spark.column
             that = other.spark.column
@@ -4977,7 +4977,7 @@ def repeat(self, repeats: Union[int, "Series"]) -> "Series":
         Series([], dtype: int64)
         """
         if not isinstance(repeats, (int, Series)):
-            raise ValueError(
+            raise TypeError(
                 "`repeats` argument must be integer or Series, but got {}".format(type(repeats))
             )