Scipy ImportError: DLL load failed: The specified module could not be found [duplicate]

Question

I have already checked the similar questions, but no one helped. I have successfully installed scipy 0.17.0 wheel provided by Christoph Gohlke on my Python 3.4 Win64 bit. When I import scipy, it works, but it does not work with the optimize, integrate, .... the other modules. I get this error

import numpy as np
from scipy import optimize

>>>  Traceback (most recent call last):
File "C:\Users\uqeirann\Desktop\test.py", line 3, in <module>
from scipy import integrate
File "C:\Python34\lib\site-packages\scipy\integrate\__init__.py", line 55, in <module>
from .quadrature import *
File "C:\Python34\lib\site-packages\scipy\integrate\quadrature.py", line 10, in <module>
from scipy.special.orthogonal import p_roots
File "C:\Python34\lib\site-packages\scipy\special\__init__.py", line 627, in <module>
from ._ufuncs import *
ImportError: DLL load failed: The specified module could not be found.

The version of my windows in 10, 64 bit and version of Python also 3.4 64 bit The version of scipy also is 64 bit. So, the problem is not related to the inconsistency, in the installed folder scipy, there are also optimize folder and the other modules. I rebooted my computer for several times, and also reinstalled scipy, didn't solve! I run pywin32, the lates version, didn't help!

I really need your prompt help

Answer 1

IME, Anaconda is the way to go if you need to get scipy etc. up & running quickly on a Windows machine

Answer 2

It's failing to import:

from ._ufuncs import *

, because it lacks the necessary Numpy dependencies. In Christoph Gohlke site you'll find the note:

SciPy is software for mathematics, science, and engineering.
       Requires numpy+mkl.

I believe if you install all necessary dependencies (probably the one I've mentioned above is enough) it should work.