Installing Scipy Error on Windows 7
In reverse, I messed up my scipy / numpy installation on my IntelX64 / Windows 7 / Python 2.7 machine. I did this while trying to play around with some neural network libraries (lasagne / theano).
To fix this, I uninstalled / reinstalled python 2.7.9 completely. After that, I went to http://www.lfd.uci.edu/~gohlke/pythonlibs/ to download and install numpy and scipy using the commands:
pip install C:\Users\me\Documents\numpy-1.9.2+mkl-cp27-none-win_amd64.whl
It went well and worked well. (I can import numpy from python)
pip install C:\Users\me\Documents\scipy-0.15.1-cp27-none-win_amd64.whl
It's also good. Except for the error in the python console:
>>> from scipy import stats
gives:
Traceback (most recent call last):
File "<stdin>", line1, in <module>
File "C:\Python27\lib\site-packages\scipy\stats\__init__.py", line 338, in <module>
from .stats import *
File "C:\Python27\lib\site-packages\scipy\stats\stats.py", line 184, in <module>
import scipy.special as special
File "C:\Python27\lib\site-packages\scipy\special\__init__.py", line 586, in <module>
from ._ufuncs import *
ImportError: DLL load failed: The specified module could not be found.
I read the required DLLs and ran the "walking walker" ( http://www.dependencywalker.com/ ) on "c: \ python27 \ lib \ site-packages \ scipy \ special_UFUNCS.PYD" and it seems there are multiple DLLs that were not found.
libifcoremd.dll
libiomp5md.dll
libmmd.dll
msvcr90.dll
dcomp.dll
ieshims.dll
and a 6 files that look like 'api-ms-win-....dll'
From my point of view, people solved this by reinstalling numpy and then scipy. I tried this and the same happens.
Does anyone else come across this problem and solve it?
source to share
I fixed this by completely removing everything (python and packages) and then installing everything again.
Just reinstalling the corrupting packages didn't work. The system also did not recover. I won't accept this answer because uninstall / install is working on a real problem. I just want to post this to help anyone else in this position.
source to share