WeichenXu123 commented on a change in pull request #24643:
[SPARK-26412][PySpark][SQL] Allow Pandas UDF to take an iterator of pd.Series
or an iterator of tuple of pd.Series
URL: https://github.com/apache/spark/pull/24643#discussion_r292981816
##########
File path: python/pyspark/worker.py
##########
@@ -255,13 +269,54 @@ def read_udfs(pickleSer, infile, eval_type):
# Scalar Pandas UDF handles struct type arguments as pandas DataFrames
instead of
# pandas Series. See SPARK-27240.
- df_for_struct = eval_type == PythonEvalType.SQL_SCALAR_PANDAS_UDF
+ df_for_struct = (eval_type == PythonEvalType.SQL_SCALAR_PANDAS_UDF or
+ eval_type ==
PythonEvalType.SQL_SCALAR_PANDAS_ITER_UDF)
ser = ArrowStreamPandasUDFSerializer(timezone, safecheck,
assign_cols_by_name,
df_for_struct)
else:
ser = BatchedSerializer(PickleSerializer(), 100)
num_udfs = read_int(infile)
+
+ if eval_type == PythonEvalType.SQL_SCALAR_PANDAS_ITER_UDF:
+ assert num_udfs == 1, "One SQL_SCALAR_PANDAS_ITER_UDF expected here."
+
+ arg_offsets, udf = read_single_udf(
+ pickleSer, infile, eval_type, runner_conf, udf_index=i)
+
+ def func(_, iterator):
+ num_input_rows = [0]
Review comment:
define `num_input_rows` outside func doesn't work.
The closure variable can only be referenced from inner function, but cannot
be assigned new value.
Another working approach is we can add a declaration `nonlocal
num_input_rows`, but python2 do not support this.
----------------------------------------------------------------
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.
For queries about this service, please contact Infrastructure at:
[email protected]
With regards,
Apache Git Services
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
