[SciPy-Dev] Fixing a bug with scipy's hypergeometric function hyp2f1

Fri Oct 27 20:26:52 EDT 2017

It might make sense to submit some of these smaller fixes as quick PRs.
They will require little review and can get merged quickly, while the
process for algorithm changes will likely take more time.

- Josh

On Fri, Oct 27, 2017 at 6:50 PM, Adam <Former at physicist.net> wrote:

> Okay *that* worked! Thank you very much!
>
> > If _generate_ufuncs.py is in your repo then you aren't working with the
> master branch of SciPy. It was changed to _generate_pyx.py recently and is
> now run automatically as part of the build process. Make sure you update
> (or there will be a ton of merge conflicts when you submit the PR).
>
>
> My repo actually does have _generate_pyx.py which is what I meant; I
> copied _generate_ufuncs.py from the comment header of add_newdocs.py.  In
> my PR, I'll change these comments so they refer to the new
> _generate_pyx.py...
>
> I'll also fix the reference to "Computation of Special Functions" in my PR
> unless someone objects...
>
> On Fri, 2017-10-27 at 16:46 -0500, Joshua Wilson wrote:
>
> > it appears I wasn't running _generate_ufuncs.py beforehand
>
> If _generate_ufuncs.py is in your repo then you aren't working with the
> master branch of SciPy. It was changed to _generate_pyx.py recently and is
> now run automatically as part of the build process. Make sure you update
> (or there will be a ton of merge conflicts when you submit the PR).
>
> > Is there a way of generating HTML that renders the latex equations?
>
> Maybe try make html-scipyorg. The first time you try it will probably
> complain that you are missing a git submodule, but it should also give
> instructions on how to fix that.
>
> > Is the 2nd author of the first reference misspelled
>
> Looks like it.
>
> On Fri, Oct 27, 2017 at 3:56 PM, Adam <Former at physicist.net> wrote:
>
> I figured it out and it appears I wasn't running _generate_ufuncs.py
> beforehand...
>
> That said, the generated HTML still doesn't render my equations as a latex
> equation.  That is to say, my output looks like:
>
>
> *Is there a way of generating HTML that renders the latex equations?*
> Like it appears on the https://docs.scipy.org/doc
> /scipy/reference/generated/scipy.special.hyp2f1.html
>
> *Bonus question: *Is the 2nd author of the first reference misspelled,
> i.e., Jjie [sic]?  Shouldn't the entire reference be:
> S. Zhang, J. Jin "Computation of Special Functions", Wiley, 1996? See
> <https://books.google.com/books/about/Computation_of_special_functions.html?id=ASfvAAAAMAAJ>
> https://books.google.com/books/about/Computation_of_spec
> ial_functions.html?id=ASfvAAAAMAAJ
>
> I can fix the first reference in my PR as well...
>
> --Adam
>
> On Thu, 2017-10-26 at 20:52 -0500, Joshua Wilson wrote:
>
>
> I do in fact get a bunch of html in the folders build/html, but the html
> for hyp2f1 doesn't contain my edits
>
> If I recall correctly, Sphinx doesn't detect changes in the `_ufuncs`
> extension module. Maybe try forcing a complete rebuild with `make
> SPHINXOPTS=-Ea html`.
>
> On Thu, Oct 26, 2017 at 8:26 PM, Adam <Former at physicist.net> wrote:
>
>
> Hey guys,
>
> How do I view the html version of a functions docstring?  I've added my
> Buhring reference to the docstring and I want to see how it renders in
> HTML.
>
> I've tried both methods described in
> https://github.com/scipy/scipy/blob/master/doc/README.txt but neither of
> them worked.
>
> When I try:
> (scipy-dev) adam at adam-P7xxDM-G:~/scipy-dev/scipy$ python setup.py
> build_sphinx
> I get:
> (...blah...blah...)
> usage: setup.py [global_opts] cmd1 [cmd1_opts] [cmd2 [cmd2_opts] ...]
>  or: setup.py --help [cmd1 cmd2 ...]
>  or: setup.py --help-commands
>  or: setup.py cmd --help
> error: invalid command 'build_sphinx'
>
> When I try :
> (scipy-dev) adam at adam-P7xxDM-G:~/scipy-dev/scipy/doc$ make html
>
> I do in fact get a bunch of html in the folders build/html, but the html
> for
> hyp2f1 doesn't contain my edits.  I've tried building html after
> reinstalling my dev version of scipy, but that didn't change anything.
>
> Thanks for any help!
>
> --Adam
>
> On Fri, 2017-10-20 at 18:23 -0700, Adam wrote:
>
> I can certainly add the formulas to the doc string if that's what people
> want.  My only concern is that the a-b=m case involves about a page of
> formula's and would take a lot of space in the doc string (I've attached
> the
> pdf from my latex file of the formulas). Originally I was thinking that the
> docstring would just contain some reference to the pdf document, e.g., "See
> the pdf at location X for the method used when a-b is an integer."
>
> But like I said, I can put them wherever they need to go; I justed wanted
> to
> make sure that future maintainers have some reference as to what the code
> is
> trying to do...
>
> --Adam
>
> On Fri, 2017-10-20 at 08:54 -0400, josef.pktd at gmail.com wrote:
>
> One possibility:
> Currently the only more extensive Latex based documentation is in the
> tutorial.
> I think it would be possible to add a technical appendix or something like
> that
> to the scipy.special tutorial, a bit similar to the distributions formulas
> attached to the
> stats tutorial.
>
> For example Boost, last time I checked, had long documentation for the
> special
> functions, which might be too long to fit in docstrings.
>
> I don't know how much there would be for special functions and whether it
> is
> difficult to maintain those notes. However, I think it would be good to
> have
> notes that developers have already written available for future
> developers and users that are interested in technical details.
>
>
> Josef
>
>
> On Fri, Oct 20, 2017 at 2:01 AM, Ralf Gommers <ralf.gommers at gmail.com>
> wrote:
>
>
>
> On Thu, Oct 19, 2017 at 8:04 PM, Ted Pudlik <tpudlik at gmail.com> wrote:
>
> Concerning the actual formulas: there's a compromise between leaving them
> implicit and creating a separate LaTeX doc.  You could add the formulas to
> the docstring.  The LaTeX will render in the HTML version of the
> documentation (example).  I'm not sure how the maintainers feel about this,
> though (this is just a suggestion from a "private citizen").
>
>
> In general: using LaTeX can be a good idea, the one thing that has to be
> kept in mind is readability as plain text (important both for reading
> docstrings in IPython terminal and when working on the code in an editor).
> Best to add LaTeX formulas to the Notes section rather than in the first
> sentences. And avoid usage of things like \left[ that make the rendered
> html
> slightly prettier but the actual math much harder to read as plain text.
>
> Here's a recent PR with discussion about various LaTeX styles:
> https://github.com/scipy/scipy/pull/7756. The style that got merged is
> about
> right.
>
> Cheers,
> Ralf
>
>
>
>
> On Thu, Oct 19, 2017 at 11:59 AM Adam <Former at physicist.net> wrote:
>
> Okay cool; thanks for the helpful reply!
>
> I'll look at Gosper's method and see how it compares with Buhring's method.
> For now I'll plan on doing a PR that implements one of these two
> methods.  I
> was just worried that I might end up doing a lot of work on a PR that
> implements Buhring's series only to have a reviewer reject it saying "Well,
> you should have used such-and-such's algorithm which is must faster, much
> more accurate, etc."
>
> I'll also hold off on adding a latex doc to the repo of the actual formulas
> used for the b-a=integer special case (unless I hear otherwise).
>
> Thanks again!
>
> --Adam
>
> -----Original Message-----
> From: Joshua Wilson
> Sent: Thursday, October 19, 2017 9:35 AM
> To: SciPy Developers List
> Subject: Re: [SciPy-Dev] Fixing a bug with scipy's hypergeometric function
> hyp2f1
>
> Hey Adam,
>
>
> Does this sound like a worthwhile PR?
>
> Yes, definitely
>
>
> Does the implementation sound reasonable?
>
> It's been a while since I've thought about this, but if I recall
> correctly the problematic region you've found is one that comes up
> quite frequently--see e.g. page 14 in
>
> http://fredrikj.net/math/hypgeom.pdf
>
> Floating around in the ether is a method credited to Bill Gosper for
> handling that region which also uses a recurrence relation (maybe
> related to/the same as in the paper you cited)? I can never seem to
> find the original reference (dead link), but I've attached someone's
> writeup of it.
>
> So, your implementation sounds reasonable, but if you really want to
> dig into it you could check out the Gosper stuff and see how they
> compare.
>
>
> Can the PR implement formulas/methods that don't appear in the literature?
> Is it going to be a problem if I implement this limit case in the PR?
>
> It's implicit in the literature, so I think it's fine.
>
>
> I don't what reference I would place hyp2f1's doc string
>
> The Buhring paper. The formula is something that an informed reader
> could figure out after reading it.
>
>
> I would be wiling to maybe add a latex doc to the PR (placed somewhere in
> the doc folder?)
>
> If I recall correctly people were opposed to adding LaTeX docs. (But
> maybe I recall incorrectly; if so please someone correct me.) I also
> have various special function write ups that might be handy for future
> devs... maybe in a separate repo?
>
> On Wed, Oct 18, 2017 at 6:47 PM, Adam <Former at physicist.net> wrote:
>
>
> Hello guys,
>
> I've found a small region in the complex plane where scipy's
> implementation
> of the hypergeometric function hyp2f1 fails. I've documented this error in
> issue 8054 on github.
>
> I am willing to submit a PR that fixes this issue. My PR would basically
> implement the analytic continuation formula given in this paper: (Buhring,
> An Analytic Continuation of the Hypergeometric Series). I've already
> implemented this series in some prototype code written in Fortran and it
> agrees well with the values returned by mpmath's implementation of hyp2f1.
>
> Before I attempt a PR, I just wanted to touch base and ask the group the
> following:
>
> 1) Does this sound like a worthwhile PR? The failure region is somewhat
> small and I don't know with what urgency people would want this fixed.
>
> 2) Does the implementation sound reasonable? My background is physics and
> so
> I haven't done a complete literature search looking for the *fastest*
> algorithm. All I can say that the Buhring's formula works and my
> implementation only seems to be about %50 slower than the current hyp2f1
> (at
> points in the complex plane where both methods converge). I would only
> apply
> Buhring's series in the region where hyp2f1 currently diverges.
>
> 3) Can the PR implement formulas/methods that don't appear in the
> literature? Buhring's paper *only* gives the analytic continuation for the
> case where the difference between the a/b parameters is NOT an integer.
> When
> a-b=m, the limit case of his series can be derived using a method
> described
> in "The Special Functions and Their Approximations" by Y. Luke (as Buhling
> mentions in his paper). I've derived the formula for this limit case and
> have an implementation of it that produces values in agreement with
> mpmath.
> Is it going to be a problem if I implement this limit case in the PR? I
> ask
> because I don't what reference I would place hyp2f1's doc string. I would
> be
> wiling to maybe add a latex doc to the PR (placed somewhere in the doc
> folder?) that contains the formula so that future scipy devs have
> something
> to reference when reviewing hyp2f1's source code.
>
> Anyways, let me know if my idea for a PR sounds like a good idea! I
> apologize for the longish email, but this is my first time trying to
> contribute to scipy...
>
> --Adam
>
> _______________________________________________
> SciPy-Dev mailing list
> SciPy-Dev at python.org
> https://mail.python.org/mailman/listinfo/scipy-dev
>
>
>
>
>
>
> _______________________________________________
> SciPy-Dev mailing list
> SciPy-Dev at python.org
> https://mail.python.org/mailman/listinfo/scipy-dev
>
> _______________________________________________
> SciPy-Dev mailing list
> SciPy-Dev at python.org
> https://mail.python.org/mailman/listinfo/scipy-dev
>
>
> _______________________________________________
> SciPy-Dev mailing list
> SciPy-Dev at python.org
> https://mail.python.org/mailman/listinfo/scipy-dev
>
>
>
> _______________________________________________
> SciPy-Dev mailing list
> SciPy-Dev at python.org
> https://mail.python.org/mailman/listinfo/scipy-dev
>
>
> _______________________________________________
> SciPy-Dev mailing list
> SciPy-Dev at python.org
> https://mail.python.org/mailman/listinfo/scipy-dev
>
> _______________________________________________
> SciPy-Dev mailing list
> SciPy-Dev at python.org
> https://mail.python.org/mailman/listinfo/scipy-dev
>
>
> _______________________________________________
> SciPy-Dev mailing list
> SciPy-Dev at python.org
> https://mail.python.org/mailman/listinfo/scipy-dev
>
> _______________________________________________
> SciPy-Dev mailing list
> SciPy-Dev at python.org
> https://mail.python.org/mailman/listinfo/scipy-dev
>
>
> _______________________________________________
> SciPy-Dev mailing list
> SciPy-Dev at python.org
> https://mail.python.org/mailman/listinfo/scipy-dev
>
>
> _______________________________________________
> SciPy-Dev mailing listSciPy-Dev at python.orghttps://mail.python.org/mailman/listinfo/scipy-dev
>
>
> _______________________________________________
> SciPy-Dev mailing list
> SciPy-Dev at python.org
> https://mail.python.org/mailman/listinfo/scipy-dev
>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.python.org/pipermail/scipy-dev/attachments/20171027/ea858204/attachment-0001.html>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: unknown-WJZT8Y
Type: image/png
Size: 86595 bytes
Desc: not available
URL: <http://mail.python.org/pipermail/scipy-dev/attachments/20171027/ea858204/attachment-0001.png>