search

pandas-dev / pandas

Flexible and powerful data analysis / manipulation library for Python, providing labeled data structures similar to R data.frame objects, statistical functions, and much more
https://pandas.pydata.org
BSD 3-Clause "New" or "Revised" License
43.49k stars 17.88k forks source link

Include .c files in repo. #1805

Closed twiecki closed 12 years ago

twiecki commented 12 years ago

pip install git+https://github.com/pydata/pandas.git

does currently not work because the .pyx files are not being cythonized (not sure why). An easy fix is to include .c files in the git repo which should make it easier for people to deploy.

I used a simple try: import cython in setup.py that cythonizes if cython is installed and uses the .c files otherwise:

https://github.com/hddm-devs/hddm/blob/develop/setup.py#L4

wesm commented 12 years ago

I don't want to include the .c files in the repo because of the diff noise. @minrk do you know what could be wrong?

minrk commented 12 years ago

I might. @twiecki - do you have vanilla setuptools or distribute?

I encountered this or similar in pyzmq, where old-style setuptools will actually explicitly convert all your .pyx extensions back to .c if pyrex is not importable. my workaround. distribute doesn't have this problem.

I can confirm on my system that a virtualenv created with setuptools cannot run the install above, but one with distribute instead can, so it's probably the same issue.

twiecki commented 12 years ago

@minrk Yes, it's in a virtualenv. Not sure what it was created with (used virtualenvwrapper), it might have both actually.

minrk commented 12 years ago

Try this, then:

pip install git+https://github.com/minrk/pandas.git@badsetuptools
minrk commented 12 years ago

(if it works, I'll do a PR)

twiecki commented 12 years ago

@minrk: yes, that did the trick

cythoning pandas/src/generated.pyx to pandas/src/generated.c

Thanks!

On Thu, Aug 23, 2012 at 2:34 PM, Min RK notifications@github.com wrote:

(if it works, I'll do a PR)

— Reply to this email directly or view it on GitHubhttps://github.com/pydata/pandas/issues/1805#issuecomment-7979354.

minrk commented 12 years ago

See #1806