GitHub user srowen opened a pull request:
https://github.com/apache/spark/pull/14948
[SPARK-17389] [MLLIB] Significant KMeans speedup with better choice of init
steps, optimizing to remove 'runs'
## What changes were proposed in this pull request?
- Deprecate KMeans 'runs' param which was a no-op since 2.0
- optimize/simplify code now that runs = 1 always
- most significantly, choose initializationSteps = 2 as default for default
k-means|| init because the paper implies that's as optimal and is much faster.
## How was this patch tested?
Jenkins
You can merge this pull request into a Git repository by running:
$ git pull https://github.com/srowen/spark SPARK-17389
Alternatively you can review and apply these changes as the patch at:
https://github.com/apache/spark/pull/14948.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 #14948
----
commit dbed7d77927e4b5181d475d3f8afd1dac31021cf
Author: Sean Owen <[email protected]>
Date: 2016-09-03T08:23:45Z
Deprecate KMeans 'runs' param which was a no-op since 2.0;
optimize/simplify code now that runs = 1 always; most significantly, choose
initializationSteps = 2 as default for default k-means|| init because the paper
implies that's as optimal and is much faster.
----
---
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.
---
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
