xinrong-meng commented on code in PR #40988:
URL: https://github.com/apache/spark/pull/40988#discussion_r1181899926
##########
python/pyspark/sql/pandas/types.py:
##########
@@ -462,3 +467,233 @@ def _convert_dict_to_map_items(s: "PandasSeriesLike") ->
"PandasSeriesLike":
:return: pandas.Series of lists of (key, value) pairs
"""
return cast("PandasSeriesLike", s.apply(lambda d: list(d.items()) if d is
not None else None))
+
+
+def _to_corrected_pandas_type(dt: DataType) -> Optional[Any]:
+ """
+ When converting Spark SQL records to Pandas `pandas.DataFrame`, the
inferred data type
+ may be wrong. This method gets the corrected data type for Pandas if that
type may be
+ inferred incorrectly.
+ """
+ import numpy as np
+
+ if type(dt) == ByteType:
+ return np.int8
+ elif type(dt) == ShortType:
+ return np.int16
+ elif type(dt) == IntegerType:
+ return np.int32
+ elif type(dt) == LongType:
+ return np.int64
+ elif type(dt) == FloatType:
+ return np.float32
+ elif type(dt) == DoubleType:
+ return np.float64
+ elif type(dt) == BooleanType:
+ return bool
+ elif type(dt) == TimestampType:
+ return np.dtype("datetime64[ns]")
+ elif type(dt) == TimestampNTZType:
+ return np.dtype("datetime64[ns]")
+ elif type(dt) == DayTimeIntervalType:
+ return np.dtype("timedelta64[ns]")
+ else:
+ return None
+
+
+def _create_converter_to_pandas(
Review Comment:
I am wondering if PySpark/Pandas data type mappings are ever documented
publicly.
--
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: [email protected]
For queries about this service, please contact Infrastructure at:
[email protected]
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
