使用setuptools将numpy作为依赖项安装

Question

这可能是一个关于this的后续问题。

我正在使用setuptools安装我的软件包。作为依赖项，我列出了numpy。我用的是Python2.7

python setup.py install

使用这个setup.py文件：

from setuptools import setup

setup(name = "test_pack", install_requires = ["numpy"])

我最后得到的错误消息是：

ImportError: No module named numpy.distutils

为了将numpy包含为依赖项并在没有安装python-dev的情况下安装它，我需要做什么？

---

完全输出python setup.py install：

running install
running bdist_egg
running egg_info
writing requirements to test_pack.egg-info/requires.txt
writing test_pack.egg-info/PKG-INFO
writing top-level names to test_pack.egg-info/top_level.txt
writing dependency_links to test_pack.egg-info/dependency_links.txt
reading manifest file 'test_pack.egg-info/SOURCES.txt'
writing manifest file 'test_pack.egg-info/SOURCES.txt'
installing library code to build/bdist.linux-x86_64/egg
running install_lib
creating build/bdist.linux-x86_64/egg
creating build/bdist.linux-x86_64/egg/test_pack
copying build/lib/test_pack/__init__.py -> build/bdist.linux-x86_64/egg/test_pack
copying build/lib/test_pack/mod.py -> build/bdist.linux-x86_64/egg/test_pack
byte-compiling build/bdist.linux-x86_64/egg/test_pack/__init__.py to __init__.pyc
byte-compiling build/bdist.linux-x86_64/egg/test_pack/mod.py to mod.pyc
creating build/bdist.linux-x86_64/egg/EGG-INFO
copying test_pack.egg-info/PKG-INFO -> build/bdist.linux-x86_64/egg/EGG-INFO
copying test_pack.egg-info/SOURCES.txt -> build/bdist.linux-x86_64/egg/EGG-INFO
copying test_pack.egg-info/dependency_links.txt -> build/bdist.linux-x86_64/egg/EGG-INFO
copying test_pack.egg-info/requires.txt -> build/bdist.linux-x86_64/egg/EGG-INFO
copying test_pack.egg-info/top_level.txt -> build/bdist.linux-x86_64/egg/EGG-INFO
creating 'dist/test_pack-0.0.0-py2.7.egg' and adding 'build/bdist.linux-x86_64/egg' to it
removing 'build/bdist.linux-x86_64/egg' (and everything under it)
Processing test_pack-0.0.0-py2.7.egg
Copying test_pack-0.0.0-py2.7.egg to /home/woltan/local/lib/python2.7/site-packages
Adding test-pack 0.0.0 to easy-install.pth file

Installed /home/woltan/local/lib/python2.7/site-packages/test_pack-0.0.0-py2.7.egg
Processing dependencies for test-pack==0.0.0
Searching for numpy
Reading http://pypi.python.org/simple/numpy/
Reading http://numpy.scipy.org
Reading http://sourceforge.net/project/showfiles.php?group_id=1369&package_id=175103
Reading http://numeric.scipy.org
Best match: numpy 1.6.1
Downloading http://pypi.python.org/packages/source/n/numpy/numpy-1.6.1.zip#md5=462c22b8eb221c78ddd51de98fbb5979
Processing numpy-1.6.1.zip
Running numpy-1.6.1/setup.py -q bdist_egg --dist-dir /tmp/easy_install-AoFmdV/numpy-1.6.1/egg-dist-tmp-JH1j2R
non-existing path in 'numpy/distutils': 'site.cfg'
Could not locate executable g77
Found executable /opt/solstudio12.2/bin/f77
gnu: no Fortran 90 compiler found
gnu: no Fortran 90 compiler found
Found executable /opt/intel/Compiler/11.1/073/bin/intel64/ifort
Could not locate executable lf95
Could not locate executable pgf90
Could not locate executable pgf77
Found executable /opt/solstudio12.2/bin/f90
Found executable /opt/solstudio12.2/bin/f95
Could not locate executable fort
_configtest.c:1: warning: conflicting types for built-in function ‘exp’
_configtest.o: In function `main':
/tmp/easy_install-AoFmdV/numpy-1.6.1/_configtest.c:6: undefined reference to `exp'
collect2: ld returned 1 exit status
_configtest.c:1: warning: conflicting types for built-in function ‘exp’
_configtest.c:1:20: error: Python.h: No such file or directory
_configtest.o: In function `main':
/tmp/easy_install-AoFmdV/numpy-1.6.1/_configtest.c:6: undefined reference to `exp'
collect2: ld returned 1 exit status
_configtest.c:1:20: error: Python.h: No such file or directory

Answer 1

要安装numpy，setuptools将下载包并从源代码处编译它。但是，编译numpy有一些先决条件，您可以检查它here。

_configtest.c:1:20: error: Python.h: No such file or directory

此错误表示至少没有安装python dev包（如果您使用的是ubuntu/debian）。

Answer 2

除非您可以访问numpy的二进制发行版（预编译/编译），否则您必须拥有python头文件，因为它需要它们来构建numpy。这就是为什么大多数包管理器附带这些包的预编译版本。例如，您可以apt-get install python-numpy，将其链接到您的virtualenv中，并且当您尝试使用install_requires=['numpy']安装程序时，应该会看到它已经安装。

Answer 3

这是一个已知问题，跟踪于^{}。

我找到了一个解决方法：将numpy添加到setup_requires。把它放在setup_requires和install_requires中似乎对最新版本的setuptools工作得很好。

所以，你的setup.py应该看起来像

setup(
    # Your setup specific stuff here
    setup_requires=["numpy"],  # Just numpy here
    install_requires=["numpy"],  # Add any of your other dependencies here
)