Github user zhengruifeng commented on the issue:
https://github.com/apache/spark/pull/16457
I think it better to discuss in the JIRA. When we come to an agreement, I
will reopen this pr.
---
If your project is set up for it, you can reply to this email and have your
reply appear on G
Github user zhengruifeng commented on the issue:
https://github.com/apache/spark/pull/16457
Agreed. Now five algs inherit `HasWeightCol`: GLR/LoR/LiR/NB/IsotonicReg
I found that some algs use `RDD[Instance]` in `train` : GLR/LoR/LiR
```
val instances: RDD[Instance] =
Github user zhengruifeng commented on the issue:
https://github.com/apache/spark/pull/16457
@srowen OK. This is the list of algs that deals with weights:
---
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 doe
Github user jkbradley commented on the issue:
https://github.com/apache/spark/pull/16457
+1 for @sethah 's comment: Algorithms should validate input data. Some
already do:
https://github.com/apache/spark/blob/b67b35f76b684c5176dc683e7491fd01b43f4467/mllib/src/main/scala/org/apache/sp
Github user srowen commented on the issue:
https://github.com/apache/spark/pull/16457
Reasonable, yeah, though I think the usages in the existing algorithms will
all have no meaningful interpretation of negative weights. At the least this
could be more targeted to code paths where the
Github user sethah commented on the issue:
https://github.com/apache/spark/pull/16457
I tend to think the actual algorithms should handle invalid weights,
instead of adding that check into instance creation. Also, this will add
overhead each time an instance is created.
Unle
Github user AmplabJenkins commented on the issue:
https://github.com/apache/spark/pull/16457
Merged build finished. Test PASSed.
---
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
e
Github user AmplabJenkins commented on the issue:
https://github.com/apache/spark/pull/16457
Test PASSed.
Refer to this link for build results (access rights to CI server needed):
https://amplab.cs.berkeley.edu/jenkins//job/SparkPullRequestBuilder/70814/
Test PASSed.
---
Github user SparkQA commented on the issue:
https://github.com/apache/spark/pull/16457
**[Test build #70814 has
finished](https://amplab.cs.berkeley.edu/jenkins/job/SparkPullRequestBuilder/70814/testReport)**
for PR 16457 at commit
[`c99293c`](https://github.com/apache/spark/commit/c
Github user SparkQA commented on the issue:
https://github.com/apache/spark/pull/16457
**[Test build #70814 has
started](https://amplab.cs.berkeley.edu/jenkins/job/SparkPullRequestBuilder/70814/testReport)**
for PR 16457 at commit
[`c99293c`](https://github.com/apache/spark/commit/c9
Github user srowen commented on the issue:
https://github.com/apache/spark/pull/16457
The doc fixes look unrelated; let's exclude those?
There are many weight fields/args in .ml that should probably also be
checked -- would you please skim them and see if there are other obviously
11 matches
Mail list logo