2010/11/7 Phil Steitz :
> On 11/7/10 10:10 AM, Mikkel Meyer Andersen wrote:
>>
>> 2010/11/7 Phil Steitz:
>>>
>>> Switching to the right list...
>>>
>>> -
>>>>>>
>>>>>> What we need there is a good algorithm for approximating the KS
>>>>>> distribution. I have been corresponding with the author of a very
>>>>>> good
>>>>>> one
>>>>>> with a Java implementation but have thus far failed in getting consent
>>>>>> to
>>>>>> release under ASL. So at this point, I am looking for an alternative
>>>>>> good
>>>>>> algorithm to implement. All suggestions / unencumbered patches
>>>>>> welcome!
>>>>>>
>>>>>> See comments on the MATH-431 for other questions.
>>>>>>
>>>>> Just to be sure of what you mean:
>>>>> Do you want to have a two-sample Kolmogorov-Smirnov test for equality
>>>>> of distributions in addition to the Mann-Whitney? Or do you need the
>>>>> Kolmogorov-Smirnov distribution (as stated for example at
>>>>>
>>>>>
>>>>>
>>>>> http://en.wikipedia.org/wiki/Kolmogorov%E2%80%93Smirnov_test#Kolmogorov_distribution
>>>>> ) in regards to the MATH-428? Sorry, but I'm at bit confused :-).
>>>>
>>>> The goal is to implement the KS test for equality of distributions (or
>>>> homogeneity against a reference distribution). To do that we need at
>>>> least
>>>> critical values of the Kolmogorov distribution. The natural way for us
>>>> to
>>>> do that would be to implement the full distribution which would be nice
>>>> to
>>>> have in the distributions package.
>>>>
>>>> Phil
>>>
>>> Have you read "Evaluating Kolmogorov’s Distribution" by Marsaglia et
>>> al. available on http://www.jstatsoft.org/v08/i18/paper ? And do you
>>> think their approach would be the way to go?
>>>
>>> I am not sure it is best. See the comments here:
>>> http://www.iro.umontreal.ca/~lecuyer/myftp/papers/ksdist.pdf
>>>
>>> Phil
>>
>> Thanks. It looks quite thorough, indeed. Was it the Java
>> implementation you didn't get a consent to release under ASL?
>>>
> Yes. I am interested in your and others' opinions on the various algorithms
> reviewed there. Could be the Marsaglia reference above is adequate for a
> start.
I'll try to make a short comparison of the two methods ASAP.
>
> Phil
>>>
>>>>>>>>
>>>>>>>> Interesting approach for the exact algorithm for Wilcoxon. If we
>>>>>>>> stay
>>>>>>>> with this, we should ack the original author of the algorithm in the
>>>>>>>> javadoc. Looks OK to use.
>
>>>>>>>
>>>>>>> Agree - both on the approach and legal part! Does the author need to
>>>>>>> sign anything but write a mail?
>>>>>>>>
>>>>>>>> Regarding the difference from R, what I usually do in this case is
>>>>>>>> look
>>>>>>>> at the R sources to try to explain the difference. Most likely in
>>>>>>>> this
>>>>>>>> case, what is going on is they are using a different estimation
>>>>>>>> algorithm
>>>>>>>> for small n or treating ties differently. The ranking options that
>>>>>>>> we
>>>>>>>> use
>>>>>>>> were largely adapted from R, so if that is the problem, it should be
>>>>>>>> easy to
>>>>>>>> test. We need to convince ourselves that ours is better or at least
>>>>>>>> a
>>>>>>>> legitimate alternative. I will take a close look this evening, but
>>>>>>>> it
>>>>>>>> looks
>>>>>>>> like the algorithm you are using should be exact. If we can't
>>>>>>>> reconcile the
>>>>>>>> difference with R, it would be good to find a way to validate
>>>>>>>> correct
>>>>>>>> functioning of the algorithm by manufacturing reference data with
>>>>>>>> known
>>>>>>>> p.
>>>>>>>
>>>>>>> I'll try to investigate the difference, hopefully tomorrow, so that
>>>>>>> formal tests can be written and included.
>>>>>>>>
>>>>>>>>> New tests: Wilcoxon signed-rank test and Mann-Whitney U
>>>>>>>>> -------------------------------------------------------
>>>>>>>>>
>>>>>>>>> Key: MATH-431
>>>>>>>>> URL: https://issues.apache.org/jira/browse/MATH-431
>>>>>>>>> Project: Commons Math
>>>>>>>>> Issue Type: New Feature
>>>>>>>>> Reporter: Mikkel Meyer Andersen
>>>>>>>>> Assignee: Mikkel Meyer Andersen
>>>>>>>>> Priority: Minor
>>>>>>>>> Attachments: MannWhitneyUTest.java,
>>>>>>>>> MannWhitneyUTestImpl.java,
>>>>>>>>> WilcoxonSignedRankTest.java, WilcoxonSignedRankTestImpl.java
>>>>>>>>>
>>>>>>>>> Original Estimate: 4h
>>>>>>>>> Remaining Estimate: 4h
>>>>>>>>>
>>>>>>>>> Wilcoxon signed-rank test and Mann-Whitney U are commonly used
>>>>>>>>> non-parametric statistical hypothesis tests (e.g. instead of
>>>>>>>>> various
>>>>>>>>> t-tests
>>>>>>>>> when normality is not present).
>>>>>>>>
>>>>>>>> --
>>>>>>>> This message is automatically generated by JIRA.
>>>>>>>> -
>>>>>>>> You can reply to this email to add a comment to the issue online.
>>>>>>>>
>>>>>>>>
>>>>>>
>>>>>>
>>>>
>>>>
>>>
>>> ---------------------------------------------------------------------
>>> To unsubscribe, e-mail: dev-unsubscribe@commons.apache.org
>>> For additional commands, e-mail: dev-help@commons.apache.org
>>>
>>>
>>
>> ---------------------------------------------------------------------
>> To unsubscribe, e-mail: dev-unsubscribe@commons.apache.org
>> For additional commands, e-mail: dev-help@commons.apache.org
>>
>
>
> ---------------------------------------------------------------------
> To unsubscribe, e-mail: dev-unsubscribe@commons.apache.org
> For additional commands, e-mail: dev-help@commons.apache.org
>
>
---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscribe@commons.apache.org
For additional commands, e-mail: dev-help@commons.apache.org