Brian Hulette created BEAM-11627:
------------------------------------
Summary: Properly support `convert_dtype=True` in Series.apply
Key: BEAM-11627
URL: https://issues.apache.org/jira/browse/BEAM-11627
Project: Beam
Issue Type: Improvement
Components: sdk-py-core
Reporter: Brian Hulette
See
https://pandas.pydata.org/pandas-docs/stable/reference/api/pandas.Series.apply.html
convert_dtype=True indicates that pandas should observe the output and set the
dtype to something other than object if possible. We should intercept this
argument and use type inference to set the dtype. We can't rely on pandas'
inference since our implementation can't observe the entire dataset.
--
This message was sent by Atlassian Jira
(v8.3.4#803005)