Skip to content

Modify serializer. #5

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Merged
cloud-fan merged 1 commit into from
Jun 16, 2017
Merged

Modify serializer. #5

Changes from all commits
Commits
Show all changes
1 commit
Select commit
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
15 changes: 8 additions & 7 deletions python/pyspark/serializers.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -566,23 +566,24 @@ def load_stream(self, stream):
import pyarrow as pa
reader = pa.open_stream(stream)
for batch in reader:
for row_index in xrange(0, batch.num_rows):
yield [batch[col_index][row_index].as_py() for col_index in xrange(0, batch.num_columns)]
for row in (zip(*[batch[col].to_pylist() for col in range(batch.num_columns)])):
yield row

def dump_stream(self, iterator, stream):
import pyarrow as pa
# the schema is set at worker.py#read_udfs_vectorized
writer = pa.RecordBatchStreamWriter(stream, self.schema)
row_id = 0
# todo: we should append data to arrow vector directly, but I can't find the API...
column_chunks = [[] for col_index in xrange(0, len(self.schema))]
column_chunks = [[] for col_index in xrange(len(self.schema))]
appends = [column_chunks[i].append for i in xrange(len(self.schema))]
for row in iterator:
if row_id < 10000:
if len(self.schema) == 1:
column_chunks[0].append(row)
appends[0](row)
else:
for col_index in xrange(0, len(self.schema)):
column_chunks[col_index].append(row[col_index])
for col_index in xrange(len(self.schema)):
appends[col_index](row[col_index])
row_id += 1
else:
self._write(column_chunks, writer)
Expand All @@ -594,7 +595,7 @@ def dump_stream(self, iterator, stream):
def _write(self, column_chunks, writer):
import pyarrow as pa
vectors = []
for col_index in xrange(0, len(self.schema)):
for col_index in xrange(len(self.schema)):
# todo: can we reuse the vector?
vector = pa.array(column_chunks[col_index], self.schema[col_index].type)
vectors.append(vector)
Expand Down