[jira] [Commented] (MATH-878) G-Test (Log-Likelihood ratio - LLR test) in math.stat.inference

["Phil Steitz (JIRA)" <ji...@apache.org>]

    [ https://issues.apache.org/jira/browse/MATH-878?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13490276#comment-13490276 ] 

Phil Steitz commented on MATH-878:
----------------------------------

Implementation code committed in r1405620.

I made no material changes - just javadoc, making a few variables final that could be final and incorporating the MATH-885 changes (externalizing array argument checks)  I also added a few more tests.

I am still working on the TestUtils changes.  Name change there will have to wait until 4.0 if we decide to do it.  I am ambivalent, as the package name .inference is what you would end up logically adding - i.e., InferenceTestUtils - but that would be redundant.  I will add a reference to Ted's paper and other discussion in the User Guide.

I am also wondering whether it may be better to make the entropy methods public and move them to StatUtils.
                
> G-Test (Log-Likelihood ratio - LLR test) in math.stat.inference
> ---------------------------------------------------------------
>
>                 Key: MATH-878
>                 URL: https://issues.apache.org/jira/browse/MATH-878
>             Project: Commons Math
>          Issue Type: New Feature
>    Affects Versions: 3.1, 3.2, 4.0
>         Environment: Netbeans
>            Reporter: Radoslav Tsvetkov
>              Labels: features, test
>             Fix For: 3.1
>
>         Attachments: MATH-878_gTest_12102012.patch, MATH-878_gTest_15102012.patch, MATH-878_gTest_26102012.patch, vcs-diff16294.patch
>
>   Original Estimate: 24h
>  Remaining Estimate: 24h
>
> 1. Implementation of G-Test (Log-Likelihood ratio LLR test for independence and goodnes-of-fit)
> 2. Reference: http://en.wikipedia.org/wiki/G-test
> 3. Reasons-Usefulness: G-tests are tests are increasingly being used in situations where chi-squared tests were previously recommended. 
> The approximation to the theoretical chi-squared distribution for the G-test is better than for the Pearson chi-squared tests. In cases where Observed >2*Expected for some cell case, the G-test is always better than the chi-squared test.
> For testing goodness-of-fit the G-test is infinitely more efficient than the chi squared test in the sense of Bahadur, but the two tests are equally efficient in the sense of Pitman or in the sense of Hodge and Lehman. 

--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators
For more information on JIRA, see: http://www.atlassian.com/software/jira