如何在不重建的情况下将ATLAS / MKL链接到现有的Numpy。
我使用Numpy来计算大型矩阵,但发现它非常慢,因为Numpy仅使用1个核来进行计算。经过大量搜索后,我发现我的Numpy没有链接到某些优化的库,例如ATLAS / MKL。这是我的numpy配置:
>>>import numpy as np
>>>np.__config__.show()
blas_info:
libraries = ['blas']
library_dirs = ['/usr/lib']
language = f77
lapack_info:
libraries = ['lapack']
library_dirs = ['/usr/lib']
language = f77
atlas_threads_info:
NOT AVAILABLE
blas_opt_info:
libraries = ['blas']
library_dirs = ['/usr/lib']
language = f77
define_macros = [('NO_ATLAS_INFO', 1)]
atlas_blas_threads_info:
NOT AVAILABLE
openblas_info:
NOT AVAILABLE
lapack_opt_info:
libraries = ['lapack', 'blas']
library_dirs = ['/usr/lib']
language = f77
define_macros = [('NO_ATLAS_INFO', 1)]
atlas_info:
NOT AVAILABLE
lapack_mkl_info:
NOT AVAILABLE
blas_mkl_info:
NOT AVAILABLE
atlas_blas_info:
NOT AVAILABLE
mkl_info:
NOT AVAILABLE
因此,我想将ATLAS / MKL链接到Numpy。但是,我的Numpy是通过PIP安装的,所以我不想手动安装,因为我想使用最新版本。我已经做了一些搜索,但它们仅用于从头开始构建。因此,我的问题是:
有什么方法可以将ATLAS / MKL链接到Numpy,而无需重新构建?
我发现配置信息保存在Numpy安装文件夹的_ config _.py中。那么修改它可以解决我的问题吗?如果是,请告诉我如何?