ANN: mpmath 0.15 released

[Fredrik Johansson]

Hi all,

Version 0.15 of mpmath is now available on the website:
http://code.google.com/p/mpmath/

It can also be downloaded from the Python Package Index:
http://pypi.python.org/pypi/mpmath/0.15

Mpmath is a pure-Python library for arbitrary-precision floating-point
arithmetic that implements an extensive set of mathematical functions. It
can be used as a standalone library or via SymPy
(http://code.google.com/p/sympy/), and is also available as a standard
component of Sage (http://sagemath.org/). The versions in Sage
and SymPy will be updated soon.

For details about the new features in this version, see the following
blog post and the changelog:
http://fredrik-j.blogspot.com/2010/06/announcing-mpmath-015.html
http://mpmath.googlecode.com/svn/tags/0.15/CHANGES

Briefly, besides many small fixes, 0.15 includes large performance
improvements for transcendental functions, new code for computing the
nontrivial zeros of the Riemann zeta function (contributed by
Juan Arias de Reyna), and many new special functions (including
generalized 2D hypergeometric series, q-functions, and new
elliptic functions). Support for complex interval arithmetic
has also been added.

Extensive documentation is available at:
http://mpmath.googlecode.com/svn/trunk/doc/build/index.html (or
equivalently)
http://mpmath.googlecode.com/svn/tags/0.15/doc/build/index.html

Bug reports and other comments are welcome on the issue tracker at
http://code.google.com/p/mpmath/issues/list or the mpmath mailing list:
http://groups.google.com/group/mpmath

Enjoy,
Fredrik Johansson

[Ondrej Certik]

Congratulations! Also good luck for the summer. What kind of special
functions will you be working on?

Ondrej

[Fredrik Johansson]

On Sun, Jun 6, 2010 at 7:12 PM, Ondrej Certik <ond...@certik.cz> wrote:

Congratulations! Also good luck for the summer. What kind of special
functions will you be working on?

Ondrej

Thanks, and the same to you! (I read on your blog that you're going to LLNL.)

Much of the work will be of a general nature. I will write more Cython code to speed up mpmath in Sage. I will also continue to polish the hypergeometric code add more special hypergeometric functions. There are some other missing special functions to be added, but it's a rather mixed list. I may be able to take requests, in case someone's favorite special function is missing :)

I also expect to do some work on symbolic special functions in Sage.

Fredrik


(Note: replying only to the sympy list since you set this as the reply-to address)

[Ondrej Certik]

On Sun, Jun 6, 2010 at 11:35 AM, Fredrik Johansson
<fredrik....@gmail.com> wrote:
> On Sun, Jun 6, 2010 at 7:12 PM, Ondrej Certik <ond...@certik.cz> wrote:
>>
>> Congratulations! Also good luck for the summer. What kind of special
>> functions will you be working on?
>>
>> Ondrej
>
> Thanks, and the same to you! (I read on your blog that you're going to
> LLNL.)
>
> Much of the work will be of a general nature. I will write more Cython code
> to speed up mpmath in Sage. I will also continue to polish the
> hypergeometric code add more special hypergeometric functions. There are
> some other missing special functions to be added, but it's a rather mixed
> list. I may be able to take requests, in case someone's favorite special
> function is missing :)

Will it be possible to also use it outside of Sage? E.g. I guess if
you do "setup.py install" in mpmath, that it would compile the Cython
codes and install it? That'd be really cool.

I think that this summer I'll try some of your coulombic special functions.

>
> I also expect to do some work on symbolic special functions in Sage.

Cool. Keep blogging about it, I always read it. :)

Ondrej

[Aaron S. Meurer]

So what needs to be done to update the SymPy install?  It would be nice if there were a guide in the Google Code wiki on how to update mpmath, just as there is a guide on how to make a release [0].  Also, should there be an issue opened for it?

Aaron Meurer

[0] - http://code.google.com/p/sympy/wiki/NewRelease

--
You received this message because you are subscribed to the Google Groups "sympy" group.
To post to this group, send email to sy...@googlegroups.com.
To unsubscribe from this group, send email to sympy+un...@googlegroups.com.
For more options, visit this group at http://groups.google.com/group/sympy?hl=en.

[Ondrej Certik]

On Sun, Jun 6, 2010 at 7:30 PM, Aaron S. Meurer <asme...@gmail.com> wrote:
> So what needs to be done to update the SymPy install?  It would be nice if
> there were a guide in the Google Code wiki on how to update mpmath, just as
> there is a guide on how to make a release [0].  Also, should there be an
> issue opened for it?

It'd be cool to have some howto. Currently, you need to copy new
mpmath into sympy/mpmath, then use:

bin/adapt_paths.py

(just read the module docstring for more info) and then run tests and
see if everything works, usually then one has to fix some little bugs.

Ondrej

[Fredrik Johansson]

On Mon, Jun 7, 2010 at 4:18 AM, Ondrej Certik <ond...@certik.cz> wrote:

Will it be possible to also use it outside of Sage? E.g. I guess if
you do "setup.py install" in mpmath, that it would compile the Cython
codes and install it? That'd be really cool.

 
It can be done in theory. The main problem is that the Cython extension code requires MPIR and the MPIR interface code in Sage.

Fredrik

[Ondrej Certik]

How many things are needed from MPIR? Looking at
mpmath/libmp/backend.py, do you need only MPZ (=sage.Integer), or do
you need more things?

Also, where in Sage will the Cython code be? Or will it be in the
mpmath repository?

Ondrej

[Fredrik Johansson]

The Cython code is here: http://hg.sagemath.org/sage-main/file/2cffe66bd642/sage/libs/mpmath

Fredrik

[Aaron S. Meurer]

I started a page here:

http://code.google.com/p/sympy/wiki/UpdateMpmath

I would be great it Frederick or those who have had experience updating mpmath could add more details/hints.

Aaron Meurer

[Aaron S. Meurer]

I have started trying to update it. So far, one non-trivial problem that I have run into is that our Interval class uses mpmath intervals for evalf(), and the mpmath interval interface has changed in this version. Checkout my branch at http://github.com/asmeurer/sympy/tree/mpmath-0.15 and run

$ ./bin/test sets

Aaron Meurer
On Jun 7, 2010, at 12:49 AM, Ondrej Certik wrote:

[asmeurer]

OK, so I have fixed everything except for the Interval thing. Also, I
am not sure if I am supposed to copy over the docs, but if I am, I
haven't done that yet either.

Aaron Meurer

On Jun 7, 11:20 am, Aaron S. Meurer <asmeu...@gmail.com> wrote:
> I have started trying to update it.  So far, one non-trivial problem that I have run into is that our Interval class uses mpmath intervals for evalf(), and the mpmath interval interface has changed in this version.  Checkout my branch athttp://github.com/asmeurer/sympy/tree/mpmath-0.15and run

[Ondrej Certik]

On Mon, Jun 7, 2010 at 10:56 AM, asmeurer <asme...@gmail.com> wrote:
> OK, so I have fixed everything except for the Interval thing.  Also, I
> am not sure if I am supposed to copy over the docs, but if I am, I
> haven't done that yet either.

All tests pass for me, except that Interval thing. As to docs, we have them in:

doc/src/modules/mpmath/

So I would just copy it over there. (I mean unless we decide not to
ship mpmath docs with sympy anymore, but I would continue shipping
it.)

Ondrej

[Aaron S. Meurer]

OK. I pushed a commit that adds the mpmath docs. All that remains is the Interval problem, which arises from mpi(S(0), 1) failing, which in turn is because Real(S(0)) returns S.Zero instead of some Real(0.0) instance. I talked with Frederick on IRC about this, and he said that he would look into the error and see what needs to be patched, mpmath or sympy.

Also, there is now a pretty good guide on how to do this in the future at

http://code.google.com/p/sympy/wiki/UpdateMpmath

which contains notes about all the things that I had to do, so anyone who wants to do this in the future should be able to just follow the guide and do it.

Aaron Meurer

[Ondrej Certik]

On Mon, Jun 7, 2010 at 2:23 PM, Aaron S. Meurer <asme...@gmail.com> wrote:
> OK.  I pushed a commit that adds the mpmath docs.  All that remains is the Interval problem, which arises from mpi(S(0), 1) failing, which in turn is because Real(S(0)) returns S.Zero instead of some Real(0.0) instance.  I talked with Frederick on IRC about this, and he said that he would look into the error and see what needs to be patched, mpmath or sympy.
>
> Also, there is now a pretty good guide on how to do this in the future at
>
> http://code.google.com/p/sympy/wiki/UpdateMpmath
>
> which contains notes about all the things that I had to do, so anyone who wants to do this in the future should be able to just follow the guide and do it.

Thanks a lot for the work that you did with this Aaron, I really
appreciate it. After the Interval is fixed, let's get it in.

Ondrej

[Aaron S. Meurer]

I went ahead and created an issue for this:

http://code.google.com/p/sympy/issues/detail?id=1955

Aaron Meurer