I found out that it's impossible to install NumPy/SciPy via installers on Windows 64-bit, that's only possible on 32-bit. Because I need more memory than a 32-bit installation gives me, I need the 64-bit version of everything.
I tried to install everything via Pip and most things worked. But when I came to SciPy, it complained about missing a Fortran compiler. So I installed Fortran via MinGW/MSYS. But you can't install SciPy right away after that, you need to reinstall NumPy. So I tried that, but now it doesn't work anymore via Pip nor via easy_install. Both give these errors:
There are a lot of errors about LNK2019 and LNK1120,.
I get a lot of errors in the range of C: C2065,C2054,C2085,C2143`, etc. They belong together I believe.
There is no Fortran linker found, but I have no idea how to install that, can't find anything on it.
And many more errors which are already out of the visible part of my cmd-windows...
The fatal error is about LNK1120:
build\lib.win-amd64-2.7\numpy\linalg\lapack_lite.pyd : fatal error LNK1120: 7 unresolved externals error: Setup script exited with error: Command "C:\Users\me\AppData\Local\Programs\Common\Microsoft\Visual C++ for Python\9.0\VC\Bin\amd64\link.exe /DLL /nologo /INCREMENTAL:NO /LIBPATH:C:\BLAS /LIBPATH:C:\Python27\libs /LIBPATH:C:\Python27\PCbuild\amd64 /LIBPATH:build\temp.win-amd64-2.7 lapack.lib blas.lib /EXPORT:initlapack_lite build\temp.win-amd64-2.7\Release\numpy\linalg\lapack_litemodule.obj /OUT:build\lib.win-amd64-2.7\numpy\linalg\lapack_lite.pyd /IMPLIB:build\temp.win-amd64-2.7\Release\numpy\linalg\lapack_lite.lib /MANIFESTFILE:build\temp.win-amd64-2.7\Release\numpy\linalg\lapack_lite.pyd.manifest" failed with exit status 1120
What is the correct way to install the 64-bit versions NumPy and SciPy on a 64-bit Windows machine? Did I miss anything? Do I need to specify something somewhere? There is no information for Windows on these problems that I can find, only for Linux or Mac OS X, but they don't help me as I can't use their commands.
回答1:
Numpy (as also some other packages like Scipy, Pandas etc.) includes lot's of C-, Cython, and Fortran code that needs to be compiled properly, before you can use it. This is, btw, also the reason why these Python-packages provide such fast Linear Algebra.
To get precompiled packages for Windows, have a look at Gohlke's Unofficial Windows Binaries or use a distribution like Winpython (just works) or Anaconda (more complex) which provide an entire preconfigured environment with lots of packages from the scientific python stack.
回答2:
You can install scipy and numpy using their wheels.
First install wheel package if it's already not there...
Example: if you're running python3.5 32 bit on Windows choose scipy-0.18.1-cp35-cp35m-win_amd64.whl then it will automatically download.
Then go to the command line and change the directory to the downloads folder and install the above wheel using pip.
Example:
cd C:\Users\[user]\Downloads pip install scipy-0.18.1-cp35-cp35m-win_amd64.whl
回答3:
Installing with pip
You can install the numpy and scipy wheels on Windows with pip in one step if you use the appropriate link from Gohlke's Unofficial Windows Binaries (mentioned by sebix) and run the Windows command prompt as Administrator. For example, in Python 3.5, you would simply use something like this:
# numpy-1.9.3+mkl for Python 3.5 on Win AMD64 pip3.5 install http://www.lfd.uci.edu/~gohlke/pythonlibs/xmshzit7/numpy-1.9.3+mkl-cp35-none-win_amd64.whl# scipy-0.16.1 for Python 3.5 on Win AMD64 pip3.5 install http://www.lfd.uci.edu/~gohlke/pythonlibs/xmshzit7/scipy-0.16.1-cp35-none-win_amd64.whl
Intel provides pre-compiled Python modules for free in their "Intel Distribution for Python". The modules are compiled against Intel's MKL (Math Kernel Library) and thus optimized for faster performance. The package includes NumPy, SciPy, scikit-learn, pandas, matplotlib, Numba, tbb, pyDAAL, Jupyter, and others. Find more information and the download link here
回答6:
If you are on windows , you wouldn't need wheel anyway! You can directly install package by downloading the 32-bit package as win32 from this link [http://www.lfd.uci.edu/~gohlke/pythonlibs/#numpy] and then move that downloaded package to cmd's current directory and open cmd and write following codepip install numpy-1.13.1+mkl-cp36-cp36m-win32.whl then do it same for scipy
For 64-bit you need to install mingw-w64 as it is gcc and compiles numpy and scipy as precompiled status.
Currently it works fine with 32-bit.So I had opted for win32 package both for numpy+mkl and scipy in that link.
Hope This works! Give a try
回答7:
You can download the needed packages from here and use pip install "Abc.whl" from the directory where you have downloaded the file.
Note that you must place command prompt in the folder where you put the .whl files after downloading them, and you must run it as administrator, worked for me on Windows 10 64-bit now python is up and running.
回答9:
Look into python wheels to solve your problem. The best part of python wheels is that they let you install C extensions with no compilers. I just installed numpy and scipy using pip in a clean python install and they both worked fine.
回答10:
for python 3.6, the following worked for me launch cmd.exe as administrator
You have to download the file with whl extension. After that, you will copy the file into your python directory then run the following command: py -3.6 -m pip install matplotlib-2.1.0-cp36-cp36m-win_amd64.whl
翻译