[
https://issues.apache.org/jira/browse/MAHOUT-674?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
]
Stanley Xu updated MAHOUT-674:
------------------------------
Attachment: trainlogsiticexample.diff
A one line patch should fix this issue and help the users could try using
category feature for simple LR training as an example.
> Let the TrainLogistic class could print the category feature weight probably
> ----------------------------------------------------------------------------
>
> Key: MAHOUT-674
> URL: https://issues.apache.org/jira/browse/MAHOUT-674
> Project: Mahout
> Issue Type: Improvement
> Components: Examples
> Affects Versions: 0.4, 0.5
> Environment: Mac OS X 10.6.7
> Reporter: Stanley Xu
> Priority: Trivial
> Labels: patch
> Fix For: 0.5
>
> Attachments: trainlogsiticexample.diff
>
> Original Estimate: 5m
> Remaining Estimate: 5m
>
> In the example of trainlogistic, if we add a category feature. It will throw
> a NullPointerException while printing out the model.
> The root cause is for category feature, the code try to use the feature name
> as the predictor rather than feature=value as the predictor while get the
> weight from the trained LR model.
--
This message is automatically generated by JIRA.
For more information on JIRA, see: http://www.atlassian.com/software/jira
