HyukjinKwon commented on code in PR #42715:
URL: https://github.com/apache/spark/pull/42715#discussion_r1308203490
##########
python/pyspark/sql/connect/dataframe.py:
##########
@@ -2005,6 +2001,32 @@ def mapInArrow(
mapInArrow.__doc__ = PySparkDataFrame.mapInArrow.__doc__
+ def foreachPartition(self, f: Callable[[Iterator[Row]], None]) -> None:
+ assert self._plan is not None
+
+ schema = self.schema
+ field_converters = [
+ ArrowTableToRowsConversion._create_converter(f.dataType) for f in
schema.fields
+ ]
+
+ def foreach_partition_func(itr: Iterable[pa.RecordBatch]) ->
Iterable[pa.RecordBatch]:
+ def flatten() -> Iterator[Row]:
+ for table in itr:
+ columnar_data = [column.to_pylist() for column in
table.columns]
+ for i in range(0, table.num_rows):
+ values = [
+ field_converters[j](columnar_data[j][i])
+ for j in range(table.num_columns)
+ ]
+ yield _create_row(fields=schema.fieldNames(),
values=values)
+
+ f(flatten())
+ return iter([])
+
+ self.mapInArrow(foreach_partition_func, schema=StructType()).collect()
Review Comment:
Using `collect` instead in case this plan gets optimized, and Python
functions are pruned out.
--
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]
