Skip to content

Conversation

@yanboliang
Copy link
Contributor

Support FPGrowth algorithm in Python API.
Should we remove "Experimental" which were marked for FPGrowth and FPGrowthModel in Scala? @jkbradley

@SparkQA
Copy link

SparkQA commented Mar 26, 2015

Test build #29237 has started for PR 5213 at commit 4f26944.

  • This patch merges cleanly.

@SparkQA
Copy link

SparkQA commented Mar 26, 2015

Test build #29237 has finished for PR 5213 at commit 4f26944.

  • This patch fails to build.
  • This patch merges cleanly.
  • This patch adds the following public classes (experimental):
    • class FPGrowthModel(JavaModelWrapper):
    • class FPGrowth(object):

@AmplabJenkins
Copy link

Test FAILed.
Refer to this link for build results (access rights to CI server needed):
https://amplab.cs.berkeley.edu/jenkins//job/SparkPullRequestBuilder/29237/
Test FAILed.

@SparkQA
Copy link

SparkQA commented Mar 26, 2015

Test build #29240 has started for PR 5213 at commit 546494a.

  • This patch merges cleanly.

@SparkQA
Copy link

SparkQA commented Mar 26, 2015

Test build #29240 has finished for PR 5213 at commit 546494a.

  • This patch passes all tests.
  • This patch merges cleanly.
  • This patch adds the following public classes (experimental):
    • class FPGrowthModel(JavaModelWrapper):
    • class FPGrowth(object):

@AmplabJenkins
Copy link

Test PASSed.
Refer to this link for build results (access rights to CI server needed):
https://amplab.cs.berkeley.edu/jenkins//job/SparkPullRequestBuilder/29240/
Test PASSed.

Copy link
Member

Choose a reason for hiding this comment

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

typo: "helper"

@jkbradley
Copy link
Member

Let's keep it Experimental for now; we can hopefully remove that tag before the 1.4 release if no issues come up before then.

Also, can you please add doc to match the Scaladoc? (We've been lazy about this with Python but should be better about making the docs match.)

Please edit python/docs/pyspark.mllib.rst to generate docs for Python. I'd follow the "pyspark.mllib.recommendation module" for settings.

Thanks!

@SparkQA
Copy link

SparkQA commented Mar 29, 2015

Test build #29364 has started for PR 5213 at commit e3f17cb.

  • This patch does not merge cleanly.

@SparkQA
Copy link

SparkQA commented Mar 29, 2015

Test build #29364 has finished for PR 5213 at commit e3f17cb.

  • This patch passes all tests.
  • This patch does not merge cleanly.
  • This patch adds the following public classes (experimental):
    • class FPGrowthModel(JavaModelWrapper):
    • class FPGrowth(object):

@AmplabJenkins
Copy link

Test PASSed.
Refer to this link for build results (access rights to CI server needed):
https://amplab.cs.berkeley.edu/jenkins//job/SparkPullRequestBuilder/29364/
Test PASSed.

@jkbradley
Copy link
Member

@yanboliang Thanks for the updates. Can you please fix the merge issues? (Rebasing off of the current master is often easiest.)

Also, can you please add documentation to FPGrowth.train()? Copying algorithm + parameter documentation from the Scala docs should be fine. That should be it.

@SparkQA
Copy link

SparkQA commented Mar 31, 2015

Test build #29463 has started for PR 5213 at commit a2d7cf7.

@SparkQA
Copy link

SparkQA commented Mar 31, 2015

Test build #29463 has finished for PR 5213 at commit a2d7cf7.

  • This patch passes all tests.
  • This patch merges cleanly.
  • This patch adds the following public classes (experimental):
    • class FPGrowthModel(JavaModelWrapper):
    • class FPGrowth(object):
  • This patch does not change any dependencies.

@AmplabJenkins
Copy link

Test PASSed.
Refer to this link for build results (access rights to CI server needed):
https://amplab.cs.berkeley.edu/jenkins//job/SparkPullRequestBuilder/29463/
Test PASSed.

Copy link
Contributor

Choose a reason for hiding this comment

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

It might be simpler to move this out of class PythonMLLibAPI to use with py4j. See #5243.

Copy link
Contributor

Choose a reason for hiding this comment

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

Empty line before this line and doc are needed. It might be convenient if we follow the Java/Scala implementation and use a namedtuple to wrap the result. So users can call items and freq instead of [0] and [1].

@AmplabJenkins
Copy link

Test PASSed.
Refer to this link for build results (access rights to CI server needed):
https://amplab.cs.berkeley.edu/jenkins//job/SparkPullRequestBuilder/29713/
Test PASSed.

Copy link
Contributor

Choose a reason for hiding this comment

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

In Python doc, we limit the line width to 72 (following PEP8). This doesn't include the code example in the doc. Please update the doc strings in your PR.

@mengxr
Copy link
Contributor

mengxr commented Apr 5, 2015

The implementation looks good to me. There are some minor issues about the docstring style. Please fix it and it should be good to go. Thanks!

@SparkQA
Copy link

SparkQA commented Apr 9, 2015

Test build #29941 has started for PR 5213 at commit 8ce0359.

@AmplabJenkins
Copy link

Test FAILed.
Refer to this link for build results (access rights to CI server needed):
https://amplab.cs.berkeley.edu/jenkins//job/SparkPullRequestBuilder/29941/
Test FAILed.

@shaneknapp
Copy link
Contributor

jenkins, test this please

@SparkQA
Copy link

SparkQA commented Apr 9, 2015

Test build #29945 has started for PR 5213 at commit 8ce0359.

@SparkQA
Copy link

SparkQA commented Apr 9, 2015

Test build #29945 has finished for PR 5213 at commit 8ce0359.

  • This patch fails Spark unit tests.
  • This patch merges cleanly.
  • This patch adds the following public classes (experimental):
    • class FPGrowthModel(JavaModelWrapper):
    • class FPGrowth(object):
  • This patch does not change any dependencies.

@AmplabJenkins
Copy link

Test FAILed.
Refer to this link for build results (access rights to CI server needed):
https://amplab.cs.berkeley.edu/jenkins//job/SparkPullRequestBuilder/29945/
Test FAILed.

@SparkQA
Copy link

SparkQA commented Apr 9, 2015

Test build #29952 has started for PR 5213 at commit ed62ead.

@SparkQA
Copy link

SparkQA commented Apr 9, 2015

Test build #29952 has finished for PR 5213 at commit ed62ead.

  • This patch fails Spark unit tests.
  • This patch merges cleanly.
  • This patch adds the following public classes (experimental):
    • class FPGrowthModel(JavaModelWrapper):
    • class FPGrowth(object):
  • This patch does not change any dependencies.

@AmplabJenkins
Copy link

Test FAILed.
Refer to this link for build results (access rights to CI server needed):
https://amplab.cs.berkeley.edu/jenkins//job/SparkPullRequestBuilder/29952/
Test FAILed.

@mengxr
Copy link
Contributor

mengxr commented Apr 9, 2015

test this please

@SparkQA
Copy link

SparkQA commented Apr 9, 2015

Test build #29961 has started for PR 5213 at commit ed62ead.

@SparkQA
Copy link

SparkQA commented Apr 9, 2015

Test build #29961 has finished for PR 5213 at commit ed62ead.

  • This patch fails Spark unit tests.
  • This patch merges cleanly.
  • This patch adds the following public classes (experimental):
    • class FPGrowthModel(JavaModelWrapper):
    • class FPGrowth(object):
  • This patch does not change any dependencies.

@AmplabJenkins
Copy link

Test FAILed.
Refer to this link for build results (access rights to CI server needed):
https://amplab.cs.berkeley.edu/jenkins//job/SparkPullRequestBuilder/29961/
Test FAILed.

@asfgit asfgit closed this in a0411ae Apr 9, 2015
@mengxr
Copy link
Contributor

mengxr commented Apr 9, 2015

LGTM. Merged into master. (The failed test are irrelevant.) @yanboliang Thanks! I created SPARK-6827 to wrap the records in freqItemsets with namedtuples.

@yanboliang yanboliang deleted the spark-6264 branch April 24, 2015 10:02
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.

6 participants