Wednesday, June 21, 2017

setting up your road to data science via python, numpy, scipy

so if you want to do start your journey to data science or just explore what all the fuss about and then get terrorized by math (math always hated me, even though i loved it, it was just one way love, sigh....no one loves me back if i love them..lol) and back off.
So scipy, numpy, panda and many other packages depend heavily on C, cython, fortran and that is why they are so fast too. so before you just install these packages via pip you might want to set the stage right.
install python 3.6.1 (in my case)
https://www.python.org/downloads/
install pip
https://packaging.python.org/tutorials/installing-packages/#id14
install wheel
pip install wheel (optional but you better do it)
then install specific wheels for numpy and scipy
download the relevant wheels from http://www.lfd.uci.edu/~gohlke/pythonlibs/
install as below
pip install numpy-1.13.0+mkl-cp36-cp36m-win32
pip install scipy-0.19.1-cp36-cp36m-win32
 
Now go to your favorite python IDE or console and run
import scipy, numpy and they both should not throw error.
Remember that if you just import numpy or scipy it might work but if you import both it might not. If you have followed this it should not be a problem to you.