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 = f77lapack_info: libraries = ['lapack'] library_dirs = ['/usr/lib'] language = f77atlas_threads_info: NOT AVAILABLEblas_opt_info: libraries = ['blas'] library_dirs = ['/usr/lib'] language = f77 define_macros = [('NO_ATLAS_INFO', 1)]atlas_blas_threads_info: NOT AVAILABLEopenblas_info: NOT AVAILABLElapack_opt_info: libraries = ['lapack', 'blas'] library_dirs = ['/usr/lib'] language = f77 define_macros = [('NO_ATLAS_INFO', 1)]atlas_info: NOT AVAILABLElapack_mkl_info: NOT AVAILABLEblas_mkl_info: NOT AVAILABLEatlas_blas_info: NOT AVAILABLEmkl_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?
添加回答
举报
0/150
提交
取消