Skip to content

Conversation

@zhengruifeng
Copy link
Contributor

@zhengruifeng zhengruifeng commented Mar 30, 2016

What changes were proposed in this pull request?

add loglikelihood in GMM.summary

How was this patch tested?

added tests

@SparkQA
Copy link

SparkQA commented Mar 30, 2016

Test build #54522 has finished for PR 12064 at commit 5e2aff7.

  • This patch passes all tests.
  • This patch merges cleanly.
  • This patch adds no public classes.

@zhengruifeng zhengruifeng changed the title [SPARK-14272][MLLIB] Evaluate GaussianMixtureModel with LogLooklihood [SPARK-14272][MLLIB] Evaluate GaussianMixtureModel with LogLikelihood Apr 10, 2016
@zhengruifeng zhengruifeng changed the title [SPARK-14272][MLLIB] Evaluate GaussianMixtureModel with LogLikelihood [SPARK-14272][ML] Evaluate GaussianMixtureModel with LogLikelihood Oct 11, 2016
@SparkQA
Copy link

SparkQA commented Oct 11, 2016

Test build #66739 has finished for PR 12064 at commit 29841d0.

  • This patch fails to build.
  • This patch merges cleanly.
  • This patch adds no public classes.

@SparkQA
Copy link

SparkQA commented Oct 11, 2016

Test build #66743 has finished for PR 12064 at commit cbe92b6.

  • This patch passes all tests.
  • This patch merges cleanly.
  • This patch adds no public classes.

@SparkQA
Copy link

SparkQA commented Oct 12, 2016

Test build #66788 has finished for PR 12064 at commit cdd829a.

  • This patch fails to build.
  • This patch merges cleanly.
  • This patch adds no public classes.

@SparkQA
Copy link

SparkQA commented Oct 12, 2016

Test build #66806 has finished for PR 12064 at commit d5b9422.

  • This patch passes all tests.
  • This patch merges cleanly.
  • This patch adds no public classes.

@SparkQA
Copy link

SparkQA commented Dec 8, 2016

Test build #69845 has finished for PR 12064 at commit 4458a5f.

  • This patch passes all tests.
  • This patch merges cleanly.
  • This patch adds no public classes.

@SparkQA
Copy link

SparkQA commented Jan 10, 2017

Test build #71113 has finished for PR 12064 at commit 8c2d529.

  • This patch fails to build.
  • This patch merges cleanly.
  • This patch adds no public classes.

@SparkQA
Copy link

SparkQA commented Jan 10, 2017

Test build #71114 has started for PR 12064 at commit 083e4f9.

@zhengruifeng
Copy link
Contributor Author

Jenkins, retest this please

@SparkQA
Copy link

SparkQA commented Jan 10, 2017

Test build #71119 has finished for PR 12064 at commit 083e4f9.

  • This patch fails Spark unit tests.
  • This patch merges cleanly.
  • This patch adds no public classes.

@SparkQA
Copy link

SparkQA commented Jan 10, 2017

Test build #71121 has finished for PR 12064 at commit 1856e59.

  • This patch passes all tests.
  • This patch merges cleanly.
  • This patch adds no public classes.

@zhengruifeng
Copy link
Contributor Author

ping @yanboliang

Copy link
Contributor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

I have a question here, should we provide the final logLikelihood of the model in its summary as well? Since lots of users will use it to evaluate the current model, that they don't need to take another pass on data.

This will expose a public API, cc @jkbradley @sethah @srowen @MLnick to discuss the API.

Copy link
Member

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

+1 for putting it in the summary. If you want to evaluate a new dataset, then let's add an evaluate() method which returns a summary.

Copy link
Contributor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

I think we don't need to bother a separate test, you can add check for logLikelihood on the existing test(multivariate data and check againt R mvnormalmixEM) which is equivalent to what you wrote but with more reasonable dataset.

@SparkQA
Copy link

SparkQA commented Jan 12, 2017

Test build #71243 has finished for PR 12064 at commit 41e1a57.

  • This patch fails Spark unit tests.
  • This patch merges cleanly.
  • This patch adds no public classes.

@SparkQA
Copy link

SparkQA commented Jan 12, 2017

Test build #71245 has finished for PR 12064 at commit 9af6c92.

  • This patch passes all tests.
  • This patch merges cleanly.
  • This patch adds no public classes.

@zhengruifeng
Copy link
Contributor Author

@yanboliang Updated! Thanks for reviewing!

@SparkQA
Copy link

SparkQA commented Jan 16, 2017

Test build #71432 has finished for PR 12064 at commit 68f72fa.

  • This patch fails Scala style tests.
  • This patch merges cleanly.
  • This patch adds no public classes.

@SparkQA
Copy link

SparkQA commented Jan 16, 2017

Test build #71436 has finished for PR 12064 at commit d333642.

  • This patch passes all tests.
  • This patch merges cleanly.
  • This patch adds no public classes.

@zhengruifeng
Copy link
Contributor Author

ping @yanboliang


@property
@since("2.2.0")
def logLikelihood(self):
Copy link
Contributor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Add this to doc test.

@SparkQA
Copy link

SparkQA commented Jan 17, 2017

Test build #71498 has finished for PR 12064 at commit 1de60b0.

  • This patch fails PySpark unit tests.
  • This patch merges cleanly.
  • This patch adds no public classes.

@SparkQA
Copy link

SparkQA commented Jan 17, 2017

Test build #71500 has finished for PR 12064 at commit d6fa8fa.

  • This patch fails PySpark unit tests.
  • This patch merges cleanly.
  • This patch adds no public classes.

@SparkQA
Copy link

SparkQA commented Jan 18, 2017

Test build #71550 has finished for PR 12064 at commit fd85c5d.

  • This patch fails Spark unit tests.
  • This patch merges cleanly.
  • This patch adds no public classes.

@zhengruifeng
Copy link
Contributor Author

jenkins, retest this please

@SparkQA
Copy link

SparkQA commented Jan 18, 2017

Test build #71566 has finished for PR 12064 at commit fd85c5d.

  • This patch fails PySpark unit tests.
  • This patch merges cleanly.
  • This patch adds no public classes.

@SparkQA
Copy link

SparkQA commented Jan 18, 2017

Test build #71573 has started for PR 12064 at commit eebae43.

@zhengruifeng
Copy link
Contributor Author

jenkins, retest this please

@SparkQA
Copy link

SparkQA commented Jan 18, 2017

Test build #71585 has finished for PR 12064 at commit eebae43.

  • This patch fails PySpark unit tests.
  • This patch merges cleanly.
  • This patch adds no public classes.

@SparkQA
Copy link

SparkQA commented Jan 18, 2017

Test build #71598 has finished for PR 12064 at commit eb27bcc.

  • This patch passes all tests.
  • This patch merges cleanly.
  • This patch adds no public classes.

@zhengruifeng zhengruifeng changed the title [SPARK-14272][ML] Evaluate GaussianMixtureModel with LogLikelihood [SPARK-14272][ML] Add Loglikelihood in GaussianMixtureSummary Jan 19, 2017
@SparkQA
Copy link

SparkQA commented Jan 19, 2017

Test build #71641 has started for PR 12064 at commit cbec946.

@yanboliang
Copy link
Contributor

Jenkins, retest this please.

@SparkQA
Copy link

SparkQA commented Jan 19, 2017

Test build #71646 has finished for PR 12064 at commit cbec946.

  • This patch passes all tests.
  • This patch merges cleanly.
  • This patch adds no public classes.

@yanboliang
Copy link
Contributor

LGTM, merged into master. Thanks!

@asfgit asfgit closed this in 8ccca91 Jan 19, 2017
@zhengruifeng zhengruifeng deleted the gmm_metric branch January 19, 2017 12:00
uzadude pushed a commit to uzadude/spark that referenced this pull request Jan 27, 2017
## What changes were proposed in this pull request?

add loglikelihood in GMM.summary

## How was this patch tested?

added tests

Author: Zheng RuiFeng <[email protected]>
Author: Ruifeng Zheng <[email protected]>

Closes apache#12064 from zhengruifeng/gmm_metric.
cmonkey pushed a commit to cmonkey/spark that referenced this pull request Feb 15, 2017
## What changes were proposed in this pull request?

add loglikelihood in GMM.summary

## How was this patch tested?

added tests

Author: Zheng RuiFeng <[email protected]>
Author: Ruifeng Zheng <[email protected]>

Closes apache#12064 from zhengruifeng/gmm_metric.
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

4 participants