If sigma is exactly an eigenvalue of A, eigs will encounter problems when it performs divisions of the form 1/(lambda - sigma), where lambda is an approximation of an eigenvalue of A. For example, eigs(A,2.0) finds the two largest magnitude eigenvalues, not the six eigenvalues closest to 2.0, as you may have wanted. If sigma is a scalar with no fractional part, k must be specified first. Computes k/2 eigenvalues from each end of the spectrum (one more from the high end if k is odd.) If sigma is one of the following strings, it specifies the desired eigenvalues: 'lm' Largest Magnitude (the default) 'sm' Smallest Magnitude (same as sigma = 0) 'lr' Largest Real part 'sr' Smallest Real part 'be' Both Ends. If sigma is a real or complex scalar, the shift, the k eigenvalues nearest sigma, are computed. ![]() If sigma is 0, the k eigenvalues smallest in magnitude are computed. If sigma is not specified, the k eigenvalues largest in magnitude are computed. If k is not specified, k = min(n,6) eigenvalues are computed.Ī scalar shift or a two letter string. If B is not specified,Īn integer, the number of eigenvalues desired. The remaining input arguments are optional and can be given in practically any order: ArgumentĪ matrix the same size as A. flag = 0 indicates convergence flag = 1 indicates no convergence. With three output arguments, flag indicates whether or not the eigenvalues were computed to the desired tolerance. ![]() ![]() With one output argument, d is a vector containing k eigenvalues.With two output arguments, V is a matrix with k columns and D is a k-by- k diagonal matrix so that A*V = V*D or A*V = B*V*D. In the latter case, the second input argument must be n, the order of the problem. Solves the eigenvalue problem where the first input argument is either a square matrix (which can be full or sparse, symmetric or nonsymmetric, real or complex), or a string containing the name of an M-file which applies a linear operator to the columns of a given matrix. Only a few selected eigenvalues, or eigenvalues and eigenvectors, are computed, in contrast to eig, which computes all eigenvalues and eigenvectors. Solves the eigenvalue problem A*v = lambda*v or the generalized eigenvalue problem A*v = lambda*B*v. Eigs (MATLAB Function Reference) MATLAB Function Reference
0 Comments
Leave a Reply. |
AuthorWrite something about yourself. No need to be fancy, just an overview. ArchivesCategories |