[jira] [Comment Edited] (MAHOUT-1493) Port Naive Bayes to the Spark DSL

["Christoph Viebig (JIRA)" <ji...@apache.org>]

    [ https://issues.apache.org/jira/browse/MAHOUT-1493?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14060102#comment-14060102 ] 

Christoph Viebig edited comment on MAHOUT-1493 at 7/13/14 1:01 PM:
-------------------------------------------------------------------

As part of an university course homework at TU Berlin, Germany, we (a group of students) got the task to continue the work done on this issue. To achieve that we ported Sebastian's code to a recent mahout master revision.
Following changes were applied to Sebastian's patch right now:
  - StandardThetaTrainer removed
  - trainNB method parameter trainComplementary removed
  - method call {{dense}} was changed to {{org.apache.mahout.math.scalabindings}} method {{dense}}
  - method call {{colSums}} was changed to {{new MatrixOps(_).colSums}}
  - method call {{rowSums}} was changed to {{new MatrixOps(_).rowSums}}
  - {{weightsPerLabelAndFeature(labelIndex, :: ) }} was changed to {{weightsPerLabelAndFeature.viewRow(labelIndex)}} as it seems to be not supported by the java class

We noticed that you refactored the code a lot in the mean time. Shall we move the patch code and the one of the used classes from mrlegacy to another location?


was (Author: cviebig):
As part of an university course homework at TU Berlin, Germany, we (a group of students) got the task to continue the work done on this issue. To achieve that we ported Sebastian's code to a recent mahout master revision.
Following changes were applied:
  - StandardThetaTrainer removed
  - trainNB method parameter trainComplementary removed
  - method call {{dense}} was changed to {{org.apache.mahout.math.scalabindings}} method {{dense}}
  - method call {{colSums}} was changed to {{new MatrixOps(_).colSums}}
  - method call {{rowSums}} was changed to {{new MatrixOps(_).rowSums}}
  - {{weightsPerLabelAndFeature(labelIndex, :: ) }} was changed to {{weightsPerLabelAndFeature.viewRow(labelIndex)}} as it seems to be not supported by the java class

We noticed that you refactored the code a lot in the mean time. Shall we move the patch code and the one of the used classes from mrlegacy to another location?

> Port Naive Bayes to the Spark DSL
> ---------------------------------
>
>                 Key: MAHOUT-1493
>                 URL: https://issues.apache.org/jira/browse/MAHOUT-1493
>             Project: Mahout
>          Issue Type: Bug
>          Components: Classification
>            Reporter: Sebastian Schelter
>            Assignee: Sebastian Schelter
>             Fix For: 1.0
>
>         Attachments: MAHOUT-1493.patch, MAHOUT-1493.patch, MAHOUT-1493.patch, MAHOUT-1493.patch
>
>
> Port our Naive Bayes implementation to the new spark dsl. Shouldn't require more than a few lines of code.



--
This message was sent by Atlassian JIRA
(v6.2#6252)