Github user srowen commented on a diff in the pull request:
https://github.com/apache/spark/pull/14858#discussion_r76571886
--- Diff: mllib/src/main/scala/org/apache/spark/ml/feature/Bucketizer.scala
---
@@ -129,17 +129,21 @@ object Bucketizer extends
DefaultParamsReadable[Bucketizer] {
if (feature == splits.last) {
splits.length - 2
} else {
- val idx = ju.Arrays.binarySearch(splits, feature)
- if (idx >= 0) {
- idx
+ if (feature.isNaN) {
--- End diff --
I think this might just be me, but, it felt like the first two checks were
at the same level:
```
if (...) {
} else if (...) {
} else {
...
}
```
Or alternatively, I could see handling the `feature == splits.last` case in
the `idx >= 0` case. Really that's the case that handle exact match of a value
and right now it's split across two clauses. Then it makes sense to start the
method by handling NaN alone.
---
If your project is set up for it, you can reply to this email and have your
reply appear on GitHub as well. If your project does not have this feature
enabled and wishes so, or if the feature is enabled but not working, please
contact infrastructure at infrastruct...@apache.org or file a JIRA ticket
with INFRA.
---
---------------------------------------------------------------------
To unsubscribe, e-mail: reviews-unsubscr...@spark.apache.org
For additional commands, e-mail: reviews-h...@spark.apache.org
