matlab编程，计算速度能提高100倍。慢的原因，是使用不对.docVIP

Loren on the Art of MATLAB March 1st, 2007 Creating Sparse Finite-Element Matrices in MATLAB Im pleased to introduce Tim Davis as this weeks guest blogger. Tim is a professor at the University of Florida, and is the author or co-author of many of our sparse matrix functions (lu, chol, much of sparse backslash, ordering methods such as amd and colamd, and other functions such as etree and symbfact). He is also the author of a recent book, Direct Methods for Sparse Linear Systems, published by SIAM, where more details of MATLAB sparse matrices are discussed ( /~davis ). Contents MATLAB is Slow? Only If You Use It Incorrectly How Not to Create a Finite-Element Matrix Whats Wrong with It? How MATLAB Stores Sparse Matrices A Better Way to Create a Finite-Element Matrix Moral: Do Not Abuse A(i,j)=... for Sparse A; Use sparse Instead Try a Faster Sparse Function When Fast is Not Fast Enough... MATLAB is Slow? Only If You Use It Incorrectly From time to time, I hear comments such as MATLAB is slow for large finite-element problems. When I look at the details, the problem is typically due to an overuse of A(i,j)= ... when creating the sparse matrix (assembling the finite elements). This can be seen in typical users code, MATLAB code in books on the topic, and even in MATLAB itself. The problem is widespread. MATLAB can be very fast for finite-element problems, but not if its used incorrectly. How Not to Create a Finite-Element Matrix A good example of what not to do can be found in the wathen.m function, in MATLAB. A=gallery(wathen,200,200) takes a huge amount of time; a very minor modification cuts the run time drastically. Im not intending to single out this one function for critique; this is a very common issue that I see over and over again. This function is built into MATLAB, which makes it an accessible example. It was first written when MATLAB did not support sparse matrices, and was modified only slightly to exploit sparsity. It was never meant for generating