How to generate a badly conditioned matrix?

17 ビュー (過去 30 日間)
Beaya
Beaya 2013 年 11 月 9 日
編集済み: Matt J 2013 年 11 月 9 日
I want to test how, badly conditioning of a matrix A, will affect calculation of A*x = b where A is my increasingly badly conditioned matrix and x is a vector of ones, so we know the exact solution. This is my code, however it gives me correct results no matter how badly conditioned my matrix is. Do you have any suggestions on what am I doing wrong?
n = 31;
for a = 1:n-1;
A = gallery('randsvd',a,a,4);
%disp(A)
x = ones(a,1);
b = A*x;
xGauss=A\b
xInv=inv(A)*b
end

回答 (1 件)

Matt J
Matt J 2013 年 11 月 9 日
編集済み: Matt J 2013 年 11 月 9 日
I don't think you've actually checked cond(A). Your matrices aren't all that ill-conditioned, for any a=1...30. Here's an example where you can see the effects of ill-conditioning
for a = 1:n-1;
A=vander(1:a);
x = ones(a,1);
b = A*x;
errorSlash=norm(A\b-1)
errorInv=norm(inv(A)*b-1)
end

カテゴリ

Help Center および File ExchangeCreating and Concatenating Matrices についてさらに検索

Translated by