GitHub user takuti opened a pull request:
https://github.com/apache/incubator-hivemall/pull/63
[HIVEMALL-90] Refine incomplete AUC UDAF implementation
## What changes were proposed in this pull request?
Since AUC UDAF (classification) did not work correctly for some specific
merge orders, this PR fixes the issue by modifying the UDAF's `merge()` and
`terminate()` implementation.
Moreover, unit tests are refined accordingly, and a utility method is
created in **HiveUtils**.
## What type of PR is it?
Bug Fix
## What is the Jira issue?
https://issues.apache.org/jira/browse/HIVEMALL-90
## How was this patch tested?
- Unit test
- Manual test on EMR
## How to use this feature?
Nothing has been changed from [current AUC
UDAF](https://hivemall.incubator.apache.org/userguide/eval/auc.html).
You can merge this pull request into a Git repository by running:
$ git pull https://github.com/takuti/incubator-hivemall fix-auc
Alternatively you can review and apply these changes as the patch at:
https://github.com/apache/incubator-hivemall/pull/63.patch
To close this pull request, make a commit to your master/trunk branch
with (at least) the following in the commit message:
This closes #63
----
commit da1578207fb9bc629455c503914757b90506ab66
Author: Takuya Kitazawa <[email protected]>
Date: 2017-03-16T02:15:01Z
Update AUC UDAF test to support all of 3!=6 merge orders
commit 5cc090fa95513dcf5db3855d5c5671cf61f45dae
Author: Takuya Kitazawa <[email protected]>
Date: 2017-03-17T03:37:05Z
Support arbitrary merge order
commit e4737fe57a555fc5719e51c1fa2881e18a44fd74
Author: Takuya Kitazawa <[email protected]>
Date: 2017-03-17T03:44:11Z
Update test case: there are two samples which have same scores
commit 5e91bbd367708d1f1e28dc00e0c64c95dfc6a66a
Author: Takuya Kitazawa <[email protected]>
Date: 2017-03-17T04:14:22Z
Fix typo
commit c0645fe74cbd0a1412747b470ead229083d03351
Author: Takuya Kitazawa <[email protected]>
Date: 2017-03-17T05:57:47Z
Carefully initialize accumulated partial area and (previous) TP/FP count
commit 627192cad19857d6a6ad92dfac18893576391053
Author: Takuya Kitazawa <[email protected]>
Date: 2017-03-17T07:14:53Z
Merge pertial result from left to right
commit a99648685a317784f5a4e2b13ed64b18cffdc4e4
Author: Takuya Kitazawa <[email protected]>
Date: 2017-03-17T14:11:15Z
Update AUC UDAF Test w/ larger sample set
commit 92b7cbc64db00a4c9994b4693a25527790ad0cee
Author: Takuya Kitazawa <[email protected]>
Date: 2017-03-17T14:13:12Z
Same scores should be passed to the same reducer
commit 49375798bdce6644b77da537b2501fcb303cb8bd
Author: Takuya Kitazawa <[email protected]>
Date: 2017-03-18T03:20:28Z
Refactor
----
---
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 [email protected] or file a JIRA ticket
with INFRA.
---
