Mailing-List: contact issues-help@commons.apache.org; run by ezmlm
Precedence: bulk
Reply-To: issues@commons.apache.org
Received-SPF: neutral (nike.apache.org: local policy)
MIME-Version: 1.0
In-Reply-To: <4CD6BB32.8030608@gmail.com>
References: <13237437.161041288507185282.JavaMail.jira@thor>
 <21457818.54431289059672301.JavaMail.jira@thor>
 <AANLkTi=+s2Em+BvcqUAQGbW+D=gX_Ys7Xnp===O02wbv@mail.gmail.com>
 <4CD6B28B.3090504@gmail.com>
 <AANLkTi=FB4TR9g2SV7yeNa=5c+1fXGaY3JsrPKB5QdZu@mail.gmail.com>
 <4CD6BB32.8030608@gmail.com>
From: Mikkel Meyer Andersen <mikl@mikl.dk>
Date: Sun, 7 Nov 2010 15:48:56 +0100
Message-ID: <AANLkTi=pOc-Gar0HVv5z_ke44cDPjP_Rd3uH2gwxY_X9@mail.gmail.com>
Subject: Re: [jira] Commented: (MATH-431) New tests: Wilcoxon signed-rank test
 and Mann-Whitney U
To: issues@commons.apache.org
Content-Type: text/plain; charset=windows-1252
Content-Transfer-Encoding: quoted-printable

2010/11/7 Phil Steitz <phil.steitz@gmail.com>:
> On 11/7/10 9:17 AM, Mikkel Meyer Andersen wrote:
>>
>> 2010/11/7 Phil Steitz<phil.steitz@gmail.com>:
>>>
>>> On 11/6/10 12:44 PM, Mikkel Meyer Andersen wrote:
>>>>
>>>> 2010/11/6 Phil Steitz (JIRA)<jira@apache.org>:
>>>>>
>>>>> =A0 =A0[
>>>>>
>>>>> https://issues.apache.org/jira/browse/MATH-431?page=3Dcom.atlassian.j=
ira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=3D129290=
54#action_12929054
>>>>> ]
>>>>>
>>>>> Phil Steitz commented on MATH-431:
>>>>> ----------------------------------
>>>>>
>>>>> +1 for including both of these tests. =A0Then on to MATH-228
>>>>
>>>> Anything I should do in regard to that?
>>>
>>> What we need there is a good algorithm for approximating the KS
>>> distribution. =A0I have been corresponding with the author of a very go=
od
>>> one
>>> with a Java implementation but have thus far failed in getting consent =
to
>>> release under ASL. =A0So at this point, I am looking for an alternative
>>> good
>>> algorithm to implement. =A0All suggestions / unencumbered patches welco=
me!
>>>
>>> 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). =A0To do that we need at l=
east
> critical values of the Kolmogorov distribution. =A0The natural way for us=
 to
> do that would be to implement the full distribution which would be nice t=
o
> have in the distributions package.
>
> Phil
Have you read "Evaluating Kolmogorov=92s 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?
>>>>>
>>>>> Interesting approach for the exact algorithm for Wilcoxon. =A0If we s=
tay
>>>>> with this, we should ack the original author of the algorithm in the
>>>>> javadoc. =A0Looks OK to use.
>>>>
>>>> Agree - both on the approach and legal part! Does the author need to
>>>> sign anything but write a mail?
>>>>>
>>>>> =A0Regarding the difference from R, what I usually do in this case is
>>>>> look
>>>>> at the R sources to try to explain the difference. =A0Most likely in =
this
>>>>> case, what is going on is they are using a different estimation
>>>>> algorithm
>>>>> for small n or treating ties differently. =A0The ranking options that=
 we
>>>>> use
>>>>> were largely adapted from R, so if that is the problem, it should be
>>>>> easy to
>>>>> test. =A0We need to convince ourselves that ours is better or at leas=
t a
>>>>> legitimate alternative. =A0I will take a close look this evening, but=
 it
>>>>> looks
>>>>> like the algorithm you are using should be exact. =A0If 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 kno=
wn
>>>>> 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
>>>>>> -------------------------------------------------------
>>>>>>
>>>>>> =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 Key: MATH-431
>>>>>> =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 URL: https://issues.apache.org/jira/=
browse/MATH-431
>>>>>> =A0 =A0 =A0 =A0 =A0 =A0 Project: Commons Math
>>>>>> =A0 =A0 =A0 =A0 =A0Issue Type: New Feature
>>>>>> =A0 =A0 =A0 =A0 =A0 =A0Reporter: Mikkel Meyer Andersen
>>>>>> =A0 =A0 =A0 =A0 =A0 =A0Assignee: Mikkel Meyer Andersen
>>>>>> =A0 =A0 =A0 =A0 =A0 =A0Priority: Minor
>>>>>> =A0 =A0 =A0 =A0 Attachments: MannWhitneyUTest.java, MannWhitneyUTest=
Impl.java,
>>>>>> WilcoxonSignedRankTest.java, WilcoxonSignedRankTestImpl.java
>>>>>>
>>>>>> =A0 Original Estimate: 4h
>>>>>> =A0Remaining 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.
>>>>>
>>>>>
>>>
>>>
>
>