mbutrovich commented on issue #1906:
URL:
https://github.com/apache/datafusion-comet/issues/1906#issuecomment-2985599126
So the big challenge here seems to be mapping Comet to Spark's execution.
Each generated Comet plan samples from its input stream, which itself is only a
single partitio
mbutrovich opened a new issue, #1906:
URL: https://github.com/apache/datafusion-comet/issues/1906
### Describe the bug
As noted by @Kontinuation, the RangePartitioning implementation that just
got merged has a bug, most obvious when trying to RangePartition pre-sorted
data:
>
