News for hackers. Stuff that matters.
Re:Python 3 in 2016 by raverbashing on Thursday February 18, 2016 @08:40AM (#11125621)
Yeah, I don't know

While Numpy/Scipy is great with a capital G, it seems the underlying infrastructure is flaky.

I'm talking about the messy process of building and installing it. And it seems everything is held with duct-tape

Also it seems that "inside Scipy' it is a different world, and that to really "talk to the outside world" there seems to be a data-translation step needed.

Re:Python 3 in 2016 by gh02t on Thursday February 18, 2016 @11:10AM (#11126735)
The complexity of building NumPy or SciPy is pretty much par for the course in HPC. Most of the libraries it is based on are battle-hardened and ridiculously well tested to a degree not seen in many places, but they date from a different era, including libraries where originally "building" it involved a hole puncher and a bucket of coffee. Getting all those libraries to play nicely together at all is a genuine miracle.

Most any significant scientific package I've ever had to build (which is quite a few) had a complex build system that felt like it was built on top of shaky glass pillars. It's just the nature of what's involved and it's a fundamentally hard problem; it's difficult to replace those libraries because the extreme amount of verification/validation and optimization that have gone into something like LAPACK makes it economically undesirable.

I'd say also that this is part of the appeal of NumPy and SciPy, which is they glue together a huge number of tools that are otherwise disjoint. You don't have to worry (most of the time) about matrix ordering, calling conventions, those mysterious "workspace" variables in LAPACK etc.

Re:Python 3 in 2016 by dr_zoidberg on Thursday February 18, 2016 @09:36AM (#11125978)
Working in Windows, my biggest issue was "the 64 bit situation", but that has been greatly solved by Microsoft's MSVS for Python (lovely move by Redmond) and the Python Wheels. Now it has truly become `pip install something` or `pip install [wheel file]` for more complex stuff.

That and Cristoph Gohlke[0]'s repository of pre-compiled modules. That guy saved me more than once. Should I ever meet him, I'll sure buy him a beer!

[0] http://www.lfd.uci.edu/~gohlke/pythonlibs/

Re:Python 3 in 2016 by ZenoArrow on Thursday February 18, 2016 @08:45AM (#11125656)
> "I'm talking about the messy process of building and installing it."

If you're using Python for scientific computing, you can use the Anaconda distribution of Python, it makes it painless to work with those packages.

https://www.continuum.io

> "Also it seems that "inside Scipy' it is a different world, and that to really "talk to the outside world" there seems to be a data-translation step needed."

Python has excellent ORMs, I'm not sure what you're getting at.

Re:Python 3 in 2016 by mangecoeur on Thursday February 18, 2016 @11:53AM (#11127111)
It's a really complicated piece of kit, that has also been around since long before modern build systems and conventions existed. That's why you have excellent scientific packages like Anaconda or Canopy, while on linux you can also install your distro's package. Situation is also becoming easier with pre-built python wheels.
Re:Python 3 in 2016 by mturmon on Thursday February 18, 2016 @01:38PM (#11128086)
Years ago, it used to be harder to install numpy, but this has changed. On MacOS, it's just

% pip install numpy

For more (e.g., other OS'es, build-from-source), see:

http://docs.scipy.org/doc/numpy-1.10.1/user/install.html

Re:Python 3 in 2016 by kmike84 on Thursday February 18, 2016 @09:40AM (#11126005)
at least on OS X installing numpy and scipy is very easy - pip installs prebuilt wheels by default, no need to compile anything