[jira] [Commented] (IGNITE-8233) KNN and SVM algorithms don't work when partition doesn't contain data

["ASF GitHub Bot (JIRA)" <ji...@apache.org>]

    [ https://issues.apache.org/jira/browse/IGNITE-8233?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16435574#comment-16435574 ] 

ASF GitHub Bot commented on IGNITE-8233:
----------------------------------------

GitHub user dmitrievanthony opened a pull request:

    https://github.com/apache/ignite/pull/3807

    IGNITE-8233 KNN and SVM algorithms don't work when partition doesn't contain data.

    

You can merge this pull request into a Git repository by running:

    $ git pull https://github.com/gridgain/apache-ignite ignite-8233

Alternatively you can review and apply these changes as the patch at:

    https://github.com/apache/ignite/pull/3807.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 #3807
    
----
commit 14a7357afc0f33b07f5d4f56d6081222ec5bb437
Author: Anton Dmitriev <dm...@...>
Date:   2018-04-12T13:35:59Z

    IGNITE-8233 Add protection of dataset compute method from empty data.

commit c63105c2cd4d4d8829ff300eaa7d11ed6620ebdb
Author: Anton Dmitriev <dm...@...>
Date:   2018-04-12T13:44:55Z

    IGNITE-8233 Fix tests after adding protection of dataset compute method
    from empty data.

----


> KNN and SVM algorithms don't work when partition doesn't contain data
> ---------------------------------------------------------------------
>
>                 Key: IGNITE-8233
>                 URL: https://issues.apache.org/jira/browse/IGNITE-8233
>             Project: Ignite
>          Issue Type: Bug
>          Components: ml
>    Affects Versions: 2.5
>            Reporter: Anton Dmitriev
>            Assignee: Anton Dmitriev
>            Priority: Major
>             Fix For: 2.5
>
>
> KNN and SVM algorithms implemented with assumption that partition data won't be null:
> {code:java}
> public LabeledDataset(double[][] mtx, double[] lbs, String[] featureNames, boolean isDistributed) {
>         super();
>         assert mtx != null;
>         assert lbs != null;
> {code}
> Currently it's wrong assumption, so we need to update dataset to support this assumption or update these algorithms.



--
This message was sent by Atlassian JIRA
(v7.6.3#76005)