ninly’s bog

More virtualenv (sans) magic

Several months back I was figuring out how to use pip and virtualenv, and generally having a pretty good time with my python development experience. However, I kept running into interesting code examples and projects that required matplotlib and the numpy/scipy stack. Since I've also been trying to get familiar with mathematical modeling, analysis and visualization techniques, these are indispensable dependencies.

I could (and had) play with them "raw", but ran into problems when I tried to enjoy the control and isolation afforded by virtualenv. Last week, though, in one of my periodic attempts to comb through dependency installations, I stumbled on this gist, and it cleared up all my problems. I haven't delved into whatever reasons may have existed for my previous difficulties, but if you're having trouble, maybe it will help you. All I did was to set up a new virtual env and pip-install my way through the list. It... just worked, for the most part (thanks, fyears!), although I did have to update some fortran-compiler stuff that numpy needed.

Pip and virtualenv sans magic

I'm trying to bridge the gap between understanding the syntax of Python (among another language or three) and how to develop in them. To wit, I was setting up some of the standard tools for package management on my Arch Linux system, and having a little trouble understanding why ...
read more

Page 1 / 1

blogroll

social