I originally posted this to the list without joining and the moderator bounced it (I know it may still be posted, so sorry for the
My security officer and sysadmins require MD5 checksums (or some other digital signature) for every piece of software that we
request for installation, and I can't find such in any of the Numpy web pages or repositories. Is that available in a place I
haven't looked or can the Numpy author(s) provide one?
David E. Sallis, Software Architect
General Dynamics Information Technology
NOAA Coastal Data Development Center
Stennis Space Center, Mississippi
"Better Living Through Software Engineering"
Just saw this come up on PyPI:
Python Bindings Generator
PyBindGen is a Python module that is geared to generating C/C++ code
that binds a C/C++ library for Python. It does so without extensive use
of either C++ templates or C pre-processor macros. It has modular
handling of C/C++ types, and can be easily extended with Python plugins.
The generated code is almost as clean as what a human programmer would
Looks interesting, esp. for C++ wrappers that Pyrex can't do.
Also, it uses Waf (http://code.google.com/p/waf/) instead of distutils,
which looks like an interesting alternative.
|David M. Cooke http://arbutus.physics.mcmaster.ca/dmc/
I have a first prototype ready for scons support in numpy.distutils
in a numpy svn branch, and would like to test it on as many
configurations as possible to see whether my design is OK. Is it
possible to add a branch to the buildbot ?
is there any more documentation about the numpy C API than the one at
This one deals mostly with creating NumPy arrays in C, but I'm more
interested in manually filling the arrays with actual data, because
wrapping of memory is not possible in my scenario.
While trying to add support for sunperf to numpy.distutils, I came
across a simple problem I am not sure how to solve best. I installed the
sun compilers for Linux, they are somewhere in my $HOME directory
($HOME/opt/sunstudio/bin). The problem is, when using the fortran
compiler, some libraries are automatically linked (the ones in
fcompiler/sun.py), which causes the build to fail by default because the
libraries are in installation path of sun compilers
Wouldn't it make sense to add the suncompiler/lib as a library path
automatically for the compiler suncompiler/bin ? If yes, how should I do
I want to generate all the possible triplets of integers in [0, n]. I am
wondering want the best possible way to do this is.
To make things clearer, I could generate i, j, k using indices:
i, j, k = indices((n, n, n))
But I will have several times the same triplet with different ordenings.
I am looking for a loop-free way of creating three arrays i, j, k with
all the triplets present once, and only once.
Any hint appreciated.
PS: I am having problems with my mail, so excuse me if this is a dup
Starting thinking over the whole distutils thing, I was thinking
what people would think about using scons inside distutils to build
extension. The more I think about it, the more I think than distutils
not being maintained, and numpy/scipy building needs being much more
complicated (at least different) than usual python extension, trying
to circumvent distutils problems is an ever ending fight. Scons, being
developped as a Make replacement, can do all we would like to be able
to do with distutils, including:
- building shared or static libraries, with dependencies (knows how
to do it on many platforms).
- can build each object file independently (e.g different compiler options)
- is much much friendlier than distutils.
- can handle external tools like swig, etc...
- have basic facility to look for libraries (ala autoconf. By basic,
I mean it is far from being as complete as autoconf, but is much
better than distutils).
Scons has also the following advantages:
- written in python, can be distributed with numpy (by this, I mean
AFAIK, license-wise, it is ok, and its size is not big): does not add
- can be called within distutils quite easily.
That is, I don't see big disadvantage to use it with distutils. It
would give use some wanted features out of the box (building
extensions based on ctypes, much friendlier way to customize building
There are some things I am not sure about :
- how to build python extension with it: this is of course mandatory
- what is required for a "bi-directional" communication with
distutils: for this to work, distutils needs to be aware of what scons
builds (for things like bdist to work, for example).
There is no question this will require some work. But anyway, my
feeling is there is a need to improve the distutils thing, and I feel
like this may be an easier path than patching over distutils
defficiencies. I know scons quite a bit, and am willing to develop at
least a prototype to see the feasibility of the whole thing.
But before starting, I would like to know whether other find the idea
attractive, dumb, is a waste of time, etc...
There is a new scipy-0.6.0.tar.gz on the sourceforge page, which
contains the missing scipy/linalg/src/fblaswrap_veclib_c.c. There is
also now a scipy-0.6.0-py2.4-win32.egg and scipy-0.6.0-py2.5-win32.egg.
Computational Infrastructure for Research Labs
10 Giannini Hall, UC Berkeley
The pure Pythonic solution is a list comprehension involving multiple sequences:
x = range(0,n)
y = x
z = x
t = [(xx,yy,zz) for xx in x for yy in y for zz in z]
You don't need subscripting, or recursive fns, or Knuth.
Runtime is almost instant (for n=10).
All the NumPy solutions look more painful, so if needs be, transform the NumPy sequences to Python ones.
Can you find the hidden words? Take a break and play Seekadoo!