[SciPy-dev] tagging 0.7rc1 this weekend?
Wed Dec 3 14:23:55 CST 2008
On Wed, Dec 3, 2008 at 3:00 PM, Pauli Virtanen <email@example.com> wrote:
> Mon, 01 Dec 2008 22:16:48 -0800, Jarrod Millman wrote:
>> I would like to get 0.7.0 out ASAP. Is there any objection to tagging a
>> release candidate this weekend? Are there any major regressions or
>> release blockers that I should be aware of?
> Maybe this should be resolved before the release (from the users list):
> Matthew Brett wrote:
>> > def ks_2samp(data1, data2):
>> > """ Computes the Kolmogorov-Smirnof statistic on 2 samples. Modified
>> > from Numerical Recipies in C, page 493. Returns KS D-value, prob. Not
>> > ufunc- like.
>> Wait - really? We can't use Numerical Recipes code, it has strict and
>> incompatible licensing... If it's in there it really has to come out as
>> fast as possible.
> According to svn log, the code for these functions originates from way
> back, r71 travo 2001-06-25 18:20:52. Apparently much of the module is
> written using NR as a reference, though I haven't checked if the code
> is similar.
> Pauli Virtanen
> Scipy-dev mailing list
There is one more case in stats.stats:
"""Calculates the t-obtained T-test on TWO RELATED samples of scores, a
and b. From Numerical Recipies, p.483. Axis can equal None (ravel array
first), or an integer (the axis over which to operate on a and b).
Returns: t-value, two-tailed p-value
In both cases it is just a few lines, and I assume the overall
algorithm changed through the adjustment to numpy/scipy
I would just drop the reference to Numerical Recipies, the algorithm
should be pretty generic.
I haven't looked in these 2 cases, but for the one-sample ks-test, in
the ticket and in the newsgroup the discussion refered to the
numerical recipes. I never looked at NR, but the proposed correction
was exactly the same as in any reference that I found on the internet,
or in articles.
More information about the Scipy-dev