maytasm opened a new pull request #11617:
URL: https://github.com/apache/druid/pull/11617
Fix bug in Variance Buffer Aggregator resulting in intermittent NaN when
druid.generic.useDefaultValueForNull=false
### Description
In aggregate method of the Buffer ObjectVarianceAggregator, we should skip
merging the ObjectVarianceAggregators when the other ObjectVarianceAggregator
has count == 0
Here is an example of how NaN can be returned:
Imagine two Buffer ObjectVarianceAggregators,
the current ObjectVarianceAggregator has count = 2, sum = 0, var =0
and other ObjectVarianceAggregator (holder2) has count =0, sum=0, var = 0
The aggregate method would run as follows:
`final double ratio = count / (double) holder2.count;`
ratio becomes infinity
then
`final double t = sum / ratio - holder2.sum;`
t becomes 0
and finally
`nvariance += holder2.nvariance + (ratio / (count + holder2.count) * t * t);`
is the same as
`holder2.nvariance + (infinity / 2 * 0 * 0);`
and infinity / 2 is NaN
Hence, variance will be NaN
This PR has:
- [x] been self-reviewed.
- [ ] using the [concurrency
checklist](https://github.com/apache/druid/blob/master/dev/code-review/concurrency.md)
(Remove this item if the PR doesn't have any relation to concurrency.)
- [ ] added documentation for new or modified features or behaviors.
- [ ] added Javadocs for most classes and all non-trivial methods. Linked
related entities via Javadoc links.
- [ ] added or updated version, license, or notice information in
[licenses.yaml](https://github.com/apache/druid/blob/master/dev/license.md)
- [ ] added comments explaining the "why" and the intent of the code
wherever would not be obvious for an unfamiliar reader.
- [x] added unit tests or modified existing tests to cover new code paths,
ensuring the threshold for [code
coverage](https://github.com/apache/druid/blob/master/dev/code-review/code-coverage.md)
is met.
- [ ] added integration tests.
- [x] been tested in a test Druid cluster.
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
To unsubscribe, e-mail: [email protected]
For queries about this service, please contact Infrastructure at:
[email protected]
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
