[SciPy-Dev] scipy.stats

[josef.pktd at gmail.com]

On Tue, Jun 1, 2010 at 4:32 AM, Travis Oliphant <oliphant at enthought.com> wrote:
>
> On Jun 1, 2010, at 3:22 AM, josef.pktd at gmail.com wrote:
>
>> On Tue, Jun 1, 2010 at 4:09 AM, Travis Oliphant <oliphant at enthought.com> wrote:
>>>
>>> On May 31, 2010, at 9:16 AM, josef.pktd at gmail.com wrote:
>>>
>>>>
>>>> This is more about the process then the content, distributions was
>>>> Travis's baby (although unfinished), and most of his changes are very
>>>> good, but I don't want to look for the 5-10% (?) typos anymore.
>>>
>>> I really am not sure what the difference between looking at timeline of changes and a formal "review" process really is?  In either case you are "looking for someone's mistakes or problems".   I do think your estimate of typos is a bit aggressive.  Really?  5-10% typos.    What is the denominator?
>>
>> I just replied for most of this.
>>
>> My test run in the middle of the weekend (before I gave up), had about
>> 4 or 5 test failures in the new _logpdf _logcdf methods.
>
> In this particular case, you can just look at the pdf method and compare it with the logpdf method.  I only added ones that were obvious.  Are you running a test different from
>
>>>> from scipy.stats import test
>>>> test()

no, I was running a variation on the new tests for logpdf logcdf, that
I have attached to the ticket

>
> to get these errors?
>
> Are you saying the skew and kurtosis test functions return different numbers than expected?

no, the methods in the distributions for distfn.stats(moments="sk") or
distfn.moment(3) or 4
I think, the f distribution is the only one where I went through the
formulas to find the typo.

I think skew and kurtosistests are ok, although I would have to look
it up to be sure.

Josef

>
> -Travis
>
>
>
> _______________________________________________
> SciPy-Dev mailing list
> SciPy-Dev at scipy.org
> http://mail.scipy.org/mailman/listinfo/scipy-dev
>