svd | 易学教程

SVD for sparse matrix in R

阅读更多关于 SVD for sparse matrix in R

问题 I've got a sparse Matrix in R that's apparently too big for me to run as.matrix() on (though it's not super-huge either). The as.matrix() call in question is inside the svd() function, so I'm wondering if anyone knows a different implementation of SVD that doesn't require first converting to a dense matrix. 回答1: The irlba package has a very fast SVD implementation for sparse matrices. 回答2: You can do a very impressive bit of sparse SVD in R using random projection as described in http://arxiv

Parallel implementation for multiple SVDs using CUDA

阅读更多关于 Parallel implementation for multiple SVDs using CUDA

问题 I'm new to parallel programming using GPU so I apologize if the question is broad or vague. I'm aware there is some parallel SVD function in the CULA library, but what should be the strategy if I have a large number of relatively small matrices to factorize? For example I have n matrices with dimension d , n is large and d is small. How to parallelize this process? Could anyone give me a hint? 回答1: My previous answer is now out-of-date. As of February 2015, CUDA 7 (currently in release

Python (NumPy, SciPy), finding the null space of a matrix

阅读更多关于 Python (NumPy, SciPy), finding the null space of a matrix

I'm trying to find the null space (solution space of Ax=0) of a given matrix. I've found two examples, but I can't seem to get either to work. Moreover, I can't understand what they're doing to get there, so I can't debug. I'm hoping someone might be able to walk me through this. The documentation pages ( numpy.linalg.svd , and numpy.compress ) are opaque to me. I learned to do this by creating the matrix C = [A|0] , finding the reduced row echelon form and solving for variables by row. I can't seem to follow how it's being done in these examples. Thanks for any and all help! Here is my sample

MATLAB eig returns inverted signs sometimes

阅读更多关于 MATLAB eig returns inverted signs sometimes

问题 I\'m trying to write a program that gets a matrix A of any size, and SVD decomposes it: A = U * S * V\' Where A is the matrix the user enters, U is an orthogonal matrix composes of the eigenvectors of A * A\' , S is a diagonal matrix of the singular values, and V is an orthogonal matrix of the eigenvectors of A\' * A . Problem is: the MATLAB function eig sometimes returns the wrong eigenvectors. This is my code: function [U,S,V]=badsvd(A) W=A*A\'; [U,S]=eig(W); max=0; for i=1:size(W,1) %%sort

Python (NumPy, SciPy), finding the null space of a matrix

阅读更多关于 Python (NumPy, SciPy), finding the null space of a matrix

问题 I\'m trying to find the null space (solution space of Ax=0) of a given matrix. I\'ve found two examples, but I can\'t seem to get either to work. Moreover, I can\'t understand what they\'re doing to get there, so I can\'t debug. I\'m hoping someone might be able to walk me through this. The documentation pages (numpy.linalg.svd, and numpy.compress) are opaque to me. I learned to do this by creating the matrix C = [A|0] , finding the reduced row echelon form and solving for variables by row. I