Github user guowei2 closed the pull request at:
https://github.com/apache/spark/pull/3821
---
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 ena
Github user guowei2 commented on the pull request:
https://github.com/apache/spark/pull/3821#issuecomment-78875126
I agree with you. Close it.
---
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
Github user marmbrus commented on the pull request:
https://github.com/apache/spark/pull/3821#issuecomment-78571134
While I would like to fix this issue, the current solution is only masking
the problem. As I said before, `java.math.BigDecimal` should never exist
during execution. I
Github user guowei2 commented on the pull request:
https://github.com/apache/spark/pull/3821#issuecomment-73838255
@rxin yeah, still exist
---
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 thi
Github user rxin commented on the pull request:
https://github.com/apache/spark/pull/3821#issuecomment-73837755
@guowei2 does this problem still exist with the latest master branch?
---
If your project is set up for it, you can reply to this email and have your
reply appear on GitHub
Github user SparkQA commented on the pull request:
https://github.com/apache/spark/pull/3821#issuecomment-70814836
[Test build #25885 has
finished](https://amplab.cs.berkeley.edu/jenkins/job/SparkPullRequestBuilder/25885/consoleFull)
for PR 3821 at commit
[`f3709f4`](https://gith
Github user AmplabJenkins commented on the pull request:
https://github.com/apache/spark/pull/3821#issuecomment-70814842
Test PASSed.
Refer to this link for build results (access rights to CI server needed):
https://amplab.cs.berkeley.edu/jenkins//job/SparkPullRequestBuilder/25
Github user SparkQA commented on the pull request:
https://github.com/apache/spark/pull/3821#issuecomment-70806372
[Test build #25885 has
started](https://amplab.cs.berkeley.edu/jenkins/job/SparkPullRequestBuilder/25885/consoleFull)
for PR 3821 at commit
[`f3709f4`](https://githu
Github user AmplabJenkins commented on the pull request:
https://github.com/apache/spark/pull/3821#issuecomment-70801024
Test FAILed.
Refer to this link for build results (access rights to CI server needed):
https://amplab.cs.berkeley.edu/jenkins//job/SparkPullRequestBuilder/25
Github user SparkQA commented on the pull request:
https://github.com/apache/spark/pull/3821#issuecomment-70801020
[Test build #25881 has
finished](https://amplab.cs.berkeley.edu/jenkins/job/SparkPullRequestBuilder/25881/consoleFull)
for PR 3821 at commit
[`a9b78d3`](https://gith
Github user SparkQA commented on the pull request:
https://github.com/apache/spark/pull/3821#issuecomment-70797203
[Test build #25881 has
started](https://amplab.cs.berkeley.edu/jenkins/job/SparkPullRequestBuilder/25881/consoleFull)
for PR 3821 at commit
[`a9b78d3`](https://githu
Github user marmbrus commented on the pull request:
https://github.com/apache/spark/pull/3821#issuecomment-69476014
I'm also still confused by this solution. The contract should be as
follows: internally we use catalyst mutable decimal type. thus, the UDF code
should only ever see
Github user SparkQA commented on the pull request:
https://github.com/apache/spark/pull/3821#issuecomment-68428335
[Test build #24944 has
finished](https://amplab.cs.berkeley.edu/jenkins/job/SparkPullRequestBuilder/24944/consoleFull)
for PR 3821 at commit
[`91db433`](https://gith
Github user AmplabJenkins commented on the pull request:
https://github.com/apache/spark/pull/3821#issuecomment-68428339
Test PASSed.
Refer to this link for build results (access rights to CI server needed):
https://amplab.cs.berkeley.edu/jenkins//job/SparkPullRequestBuilder/24
Github user guowei2 commented on the pull request:
https://github.com/apache/spark/pull/3821#issuecomment-68426263
@marmbrus I add some test cases for this.
But I'm puzzled that whether we should not change `Decimal` to `BigDecimal`
in `convertToScala`
---
If your project is set
Github user SparkQA commented on the pull request:
https://github.com/apache/spark/pull/3821#issuecomment-68426164
[Test build #24944 has
started](https://amplab.cs.berkeley.edu/jenkins/job/SparkPullRequestBuilder/24944/consoleFull)
for PR 3821 at commit
[`91db433`](https://githu
Github user guowei2 commented on the pull request:
https://github.com/apache/spark/pull/3821#issuecomment-68419415
@marmbrus i figured out it in the issue. in this case, execution
`TakeOrdered` invokes `convertRowToScala` to change `Decimal` into
`BigDecimal`.
i've tried to rem
Github user marmbrus commented on the pull request:
https://github.com/apache/spark/pull/3821#issuecomment-68388403
Thanks for reporting this issue. Can you add a test case? Also, I'm
worried this is only masking a symptom and not fixing the underlying issue.
Decimals should always
Github user SparkQA commented on the pull request:
https://github.com/apache/spark/pull/3821#issuecomment-68339547
[Test build #24887 has
finished](https://amplab.cs.berkeley.edu/jenkins/job/SparkPullRequestBuilder/24887/consoleFull)
for PR 3821 at commit
[`1bab9e4`](https://gith
Github user AmplabJenkins commented on the pull request:
https://github.com/apache/spark/pull/3821#issuecomment-68339552
Test PASSed.
Refer to this link for build results (access rights to CI server needed):
https://amplab.cs.berkeley.edu/jenkins//job/SparkPullRequestBuilder/24
Github user SparkQA commented on the pull request:
https://github.com/apache/spark/pull/3821#issuecomment-68336490
[Test build #24887 has
started](https://amplab.cs.berkeley.edu/jenkins/job/SparkPullRequestBuilder/24887/consoleFull)
for PR 3821 at commit
[`1bab9e4`](https://githu
Github user marmbrus commented on the pull request:
https://github.com/apache/spark/pull/3821#issuecomment-68336405
ok to test
---
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
ena
22 matches
Mail list logo
