大型稀疏矩阵的本征分解
问题描述：
我使用Matlab计算方阵A的维数约简，我的问题是我在计算13000 x 13000矩阵A的eigvalue分解时遇到问题，即
Im computing dimensioality reduction of a square matrix A using matlab,my issue now is that I have problem computing eigvalue decomposition of a 13000 x 13000 matrix A,ie
[v d]= eigs(A)bcos有很多零(稀疏)，我得到使用4GB RAM出现内存不足错误"，现在我确信这不是我的电脑问题，因为在运行eig(s)命令时内存没有用完.我在网上看到的帮助建议我在C ++/C中使用ARPACK lib.欢迎任何解决方案.
bcos there are a lot zeros(sparse), I get ''out of memory error'' using a 4GB RAM,now Im convinced is not my pc problem, since the memory is not used up when eig(s) command is run. the help I saw online suggested I use ARPACK lib with C++/C. Please any solution is welcome
答
几天前，我使用开源线性代数库 SCALAPACK 对于并行处理系统.我建议使用这些库.
Some days back i worked with open source linear algebra library LAPACK for single processor.Take a look on it.By the way There is another version of same library SCALAPACK for parallel processing system.I suggest these libraries.