HyukjinKwon commented on a change in pull request #34058: URL: https://github.com/apache/spark/pull/34058#discussion_r714452107
########## File path: python/pyspark/pandas/typedef/typehints.py ########## @@ -673,98 +673,146 @@ def create_tuple_for_frame_type(params: Any) -> object: Typing data columns with an index: >>> ps.DataFrame[int, [int, int]] # doctest: +ELLIPSIS - typing.Tuple[...IndexNameType, int, int] + typing.Tuple[...IndexNameType, ...NameType, ...NameType] >>> ps.DataFrame[pdf.index.dtype, pdf.dtypes] # doctest: +ELLIPSIS - typing.Tuple[...IndexNameType, numpy.int64] + typing.Tuple[...IndexNameType, ...NameType] >>> ps.DataFrame[("index", int), [("id", int), ("A", int)]] # doctest: +ELLIPSIS typing.Tuple[...IndexNameType, ...NameType, ...NameType] >>> ps.DataFrame[(pdf.index.name, pdf.index.dtype), zip(pdf.columns, pdf.dtypes)] ... # doctest: +ELLIPSIS typing.Tuple[...IndexNameType, ...NameType] + + Typing data columns with an Multi-index: + >>> arrays = [[1, 1, 2], ['red', 'blue', 'red']] + >>> idx = pd.MultiIndex.from_arrays(arrays, names=('number', 'color')) + >>> pdf = pd.DataFrame({'a': range(3)}, index=idx) + >>> ps.DataFrame[[int, int], [int, int]] # doctest: +ELLIPSIS + typing.Tuple[...IndexNameType, ...IndexNameType, ...NameType, ...NameType] + >>> ps.DataFrame[pdf.index.dtypes, pdf.dtypes] # doctest: +ELLIPSIS + typing.Tuple[...IndexNameType, ...NameType] + >>> ps.DataFrame[[("index-1", int), ("index-2", int)], [("id", int), ("A", int)]] + ... # doctest: +ELLIPSIS + typing.Tuple[...IndexNameType, ...IndexNameType, ...NameType, ...NameType] + >>> ps.DataFrame[zip(pdf.index.names, pdf.index.dtypes), zip(pdf.columns, pdf.dtypes)] + ... # doctest: +ELLIPSIS + typing.Tuple[...IndexNameType, ...NameType] """ return Tuple[extract_types(params)] # TODO(SPARK-36708): numpy.typing (numpy 1.21+) support for nested types. def extract_types(params: Any) -> Tuple: origin = params - if isinstance(params, zip): # type: ignore - # Example: - # DataFrame[zip(pdf.columns, pdf.dtypes)] - params = tuple(slice(name, tpe) for name, tpe in params) # type: ignore - if isinstance(params, Iterable): - params = tuple(params) - else: - params = (params,) + params = _prepare_a_tuple(params) - if all( - isinstance(param, slice) - and param.start is not None - and param.step is None - and param.stop is not None - for param in params - ): + if _is_valid_slices(params): # Example: # DataFrame["id": int, "A": int] - new_params = [] - for param in params: - new_param = type("NameType", (NameTypeHolder,), {}) # type: Type[NameTypeHolder] - new_param.name = param.start - # When the given argument is a numpy's dtype instance. - new_param.tpe = param.stop.type if isinstance(param.stop, np.dtype) else param.stop - new_params.append(new_param) - + new_params = _convert_slices_to_holders(params, is_index=False) return tuple(new_params) elif len(params) == 2 and isinstance(params[1], (zip, list, pd.Series)): # Example: # DataFrame[int, [int, int]] # DataFrame[pdf.index.dtype, pdf.dtypes] # DataFrame[("index", int), [("id", int), ("A", int)]] # DataFrame[(pdf.index.name, pdf.index.dtype), zip(pdf.columns, pdf.dtypes)] + # + # DataFrame[[int, int], [int, int]] + # DataFrame[pdf.index.dtypes, pdf.dtypes] Review comment: ohh okay, its for dtype*s* ########## File path: python/pyspark/pandas/typedef/typehints.py ########## @@ -673,98 +673,146 @@ def create_tuple_for_frame_type(params: Any) -> object: Typing data columns with an index: >>> ps.DataFrame[int, [int, int]] # doctest: +ELLIPSIS - typing.Tuple[...IndexNameType, int, int] + typing.Tuple[...IndexNameType, ...NameType, ...NameType] >>> ps.DataFrame[pdf.index.dtype, pdf.dtypes] # doctest: +ELLIPSIS - typing.Tuple[...IndexNameType, numpy.int64] + typing.Tuple[...IndexNameType, ...NameType] >>> ps.DataFrame[("index", int), [("id", int), ("A", int)]] # doctest: +ELLIPSIS typing.Tuple[...IndexNameType, ...NameType, ...NameType] >>> ps.DataFrame[(pdf.index.name, pdf.index.dtype), zip(pdf.columns, pdf.dtypes)] ... # doctest: +ELLIPSIS typing.Tuple[...IndexNameType, ...NameType] + + Typing data columns with an Multi-index: + >>> arrays = [[1, 1, 2], ['red', 'blue', 'red']] + >>> idx = pd.MultiIndex.from_arrays(arrays, names=('number', 'color')) + >>> pdf = pd.DataFrame({'a': range(3)}, index=idx) + >>> ps.DataFrame[[int, int], [int, int]] # doctest: +ELLIPSIS + typing.Tuple[...IndexNameType, ...IndexNameType, ...NameType, ...NameType] + >>> ps.DataFrame[pdf.index.dtypes, pdf.dtypes] # doctest: +ELLIPSIS + typing.Tuple[...IndexNameType, ...NameType] + >>> ps.DataFrame[[("index-1", int), ("index-2", int)], [("id", int), ("A", int)]] + ... # doctest: +ELLIPSIS + typing.Tuple[...IndexNameType, ...IndexNameType, ...NameType, ...NameType] + >>> ps.DataFrame[zip(pdf.index.names, pdf.index.dtypes), zip(pdf.columns, pdf.dtypes)] + ... # doctest: +ELLIPSIS + typing.Tuple[...IndexNameType, ...NameType] """ return Tuple[extract_types(params)] # TODO(SPARK-36708): numpy.typing (numpy 1.21+) support for nested types. def extract_types(params: Any) -> Tuple: origin = params - if isinstance(params, zip): # type: ignore - # Example: - # DataFrame[zip(pdf.columns, pdf.dtypes)] - params = tuple(slice(name, tpe) for name, tpe in params) # type: ignore - if isinstance(params, Iterable): - params = tuple(params) - else: - params = (params,) + params = _prepare_a_tuple(params) - if all( - isinstance(param, slice) - and param.start is not None - and param.step is None - and param.stop is not None - for param in params - ): + if _is_valid_slices(params): # Example: # DataFrame["id": int, "A": int] - new_params = [] - for param in params: - new_param = type("NameType", (NameTypeHolder,), {}) # type: Type[NameTypeHolder] - new_param.name = param.start - # When the given argument is a numpy's dtype instance. - new_param.tpe = param.stop.type if isinstance(param.stop, np.dtype) else param.stop - new_params.append(new_param) - + new_params = _convert_slices_to_holders(params, is_index=False) return tuple(new_params) elif len(params) == 2 and isinstance(params[1], (zip, list, pd.Series)): # Example: # DataFrame[int, [int, int]] # DataFrame[pdf.index.dtype, pdf.dtypes] # DataFrame[("index", int), [("id", int), ("A", int)]] # DataFrame[(pdf.index.name, pdf.index.dtype), zip(pdf.columns, pdf.dtypes)] + # + # DataFrame[[int, int], [int, int]] + # DataFrame[pdf.index.dtypes, pdf.dtypes] Review comment: ohh okay, its for dtype**s** -- This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. To unsubscribe, e-mail: reviews-unsubscr...@spark.apache.org For queries about this service, please contact Infrastructure at: us...@infra.apache.org --------------------------------------------------------------------- To unsubscribe, e-mail: reviews-unsubscr...@spark.apache.org For additional commands, e-mail: reviews-h...@spark.apache.org