[GitHub] spark pull request: SPARK-1793 - Heavily duplicated test setup cod...
Github user ajtulloch commented on the pull request: https://github.com/apache/spark/pull/726#issuecomment-43426389 Ping? --- 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 infrastruct...@apache.org or file a JIRA ticket with INFRA. ---
[GitHub] spark pull request: SPARK-1791 - SVM implementation does not use t...
GitHub user ajtulloch opened a pull request: https://github.com/apache/spark/pull/725 SPARK-1791 - SVM implementation does not use threshold parameter Summary: https://issues.apache.org/jira/browse/SPARK-1791 Simple fix, and backward compatible, since - anyone who set the threshold was getting completely wrong answers. - anyone who did not set the threshold had the default 0.0 value for the threshold anyway. Test Plan: Unit test added that is verified to fail under the old implementation, and pass under the new implementation. Reviewers: CC: You can merge this pull request into a Git repository by running: $ git pull https://github.com/ajtulloch/spark SPARK-1791-SVM Alternatively you can review and apply these changes as the patch at: https://github.com/apache/spark/pull/725.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 #725 commit 6f7075a75927b08c3c1211642632e03c2246f5cd Author: Andrew Tulloch and...@tullo.ch Date: 2014-05-10T23:22:02Z SPARK-1791 - SVM implementation does not use threshold parameter Summary: https://issues.apache.org/jira/browse/SPARK-1791 Simple fix, and backward compatible, since - anyone who set the threshold was getting completely wrong answers. - anyone who did not set the threshold had the default 0.0 value for the threshold anyway. Test Plan: Unit test added that is verified to fail under the old implementation, and pass under the new implementation. Reviewers: CC: --- 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 infrastruct...@apache.org or file a JIRA ticket with INFRA. ---
[GitHub] spark pull request: SPARK-1793 - Heavily duplicated test setup cod...
Github user ajtulloch commented on the pull request: https://github.com/apache/spark/pull/726#issuecomment-42758353 `git diff --patience` gives a cleaner view - https://gist.github.com/ajtulloch/3d9925c2f2252283ef61 --- 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 infrastruct...@apache.org or file a JIRA ticket with INFRA. ---
[GitHub] spark pull request: [SPARK-1543][MLlib] Add ADMM for solving Lasso...
Github user ajtulloch commented on the pull request: https://github.com/apache/spark/pull/458#issuecomment-42758859 I agree with @etrain, it's possible to abstract out the ADMM optimisation routine such that it's trivial to implement L1-logistic regression, lasso, [SVMs](http://web.eecs.umich.edu/~honglak/aistats12-admmDistributedSVM.pdf), etc with very few additional lines of code. I implemented that for Spark a few months ago (albeit a naive, unperformant, and untested implementation). If you wanted to see an alternative way of structuring this diff, my code is available at [ajtulloch/spark/SPARK-1794-GenericADMM](https://github.com/ajtulloch/spark/tree/SPARK-1794-GenericADMM). --- 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 infrastruct...@apache.org or file a JIRA ticket with INFRA. ---