FEAT: Support Spark filter with window operation #2687
Merged
Conversation
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
icexelloss
reviewed
Mar 17, 2021
icexelloss
reviewed
Mar 17, 2021
icexelloss
reviewed
Mar 17, 2021
icexelloss
reviewed
Mar 17, 2021
icexelloss
reviewed
Mar 22, 2021
emilyreff7
force-pushed
the
fix-filter-with-window-spark
branch
from
ea54033
to
c2366e0
Compare
|
@emilyreff7 can you merge master one more time |
emilyreff7
force-pushed
the
fix-filter-with-window-spark
branch
from
c2366e0
to
7dc1384
Compare
|
@emilyreff can you rebase |
emilyreff7
force-pushed
the
fix-filter-with-window-spark
branch
from
5906212
to
5e1b5f6
Compare
icexelloss
reviewed
Mar 23, 2021
icexelloss
changed the title
emilyreff7
force-pushed
the
fix-filter-with-window-spark
branch
2 times, most recently
from
6eb130c
to
89b9000
Compare
|
lgtm can you add a whatsnew note :-> ping on green. |
icexelloss
reviewed
Mar 23, 2021
jreback
approved these changes
Mar 23, 2021
|
@icexelloss merge when ready |
|
@jreback Looks good to me. |
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
Overview
Currently, if you write an expression that does a filter on a window operation and execute this in the pyspark backend, we get a failure from upstream Spark. For example:
table.filter(lambda t: t['id'].mean().over(window) > 3)would raise
pyspark.sql.utils.AnalysisException: It is not allowed to use window functions inside WHERE clause;. There is an upstream Spark issue about this: https://issues.apache.org/jira/browse/SPARK-33057.Proposed Change
A simple workaround can be implemented to avoid passing a window operation directly to filter in the pyspark backend. We can achieve the same filter by first assigning the filter predicate to a temporary, internal column, performing the filter using that column, and then dropping the column.
Tests
Added a test in test_basic.py that does a window operation in a filter.