Fix arrays_zip to not rely on broken segmented gather #7484
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
Signed-off-by: Robert (Bobby) Evans <[email protected]>
|
build |
|
I filed #7485 as follow on work for this. |
|
I started to read the code in this PR and wonder if this is a bug or something we need to change in |
|
@ttnghia you have it backwards. The old version would not return a zero length segment for NULL. Your patch to make_list rapidsai/cudf#12370 fixed it and now it does the right thing and produce a zero length segment. I don't know if it is guaranteed to always do that or if rapidsai/cudf#12370 is fixing it up afterwards. Either way this patch is dealing with that new reality. |
jlowe
reviewed
Jan 10, 2023
sql-plugin/src/main/scala/org/apache/spark/sql/rapids/collectionOperations.scala
Outdated
Show resolved
Hide resolved
sql-plugin/src/main/scala/org/apache/spark/sql/rapids/collectionOperations.scala
Outdated
Show resolved
Hide resolved
sql-plugin/src/main/scala/org/apache/spark/sql/rapids/collectionOperations.scala
Outdated
Show resolved
Hide resolved
sql-plugin/src/main/scala/org/apache/spark/sql/rapids/collectionOperations.scala
Outdated
Show resolved
Hide resolved
|
build |
jlowe
approved these changes
Jan 10, 2023
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Outdated suggestions cannot be applied.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.
This fixes #7469
Sorry that the patch is so large. I had to fundamentally change how some parts of this work and then did some refactoring to clean it up.
Before the code relied on segmented gather returning a child value for every gather entry even if the input column was null. then they would hack a null back on the output, possibly creating an invalid output where a row was marked as NULL but had a non-null offset.
This changed it so that we clean up the inputs so that the validity is the same everywhere. This should cause the children output from the segmented gather to all have the exact same offsets, so we can pull them apart properly. This is inefficient, because we have to make a copy of all the input if any of them have a NULL. But I didn't want to risk passing an invalid column to segmented gather and having it produce an incorrect result.
I plan on filing another issue for us to stop using
mergeAndSetValidityon any nested type. This includes strings.