I got this Error :Matrix dimensions must agree
1 次查看(过去 30 天)
显示 更早的评论
Q1: When i input this formula ; f=(4/pi)*sum((1./n)*sin(2*pi*n.*t)) I got this error: Matrix dimensions must agree
Where n=1:2:401 and t=linspace(-0.5,0.5,250)
Q2: f=1/pi+0.5*sin(pi*tau)-2/pi*(cos(pi*n'*tau)/n'.^2-1) tau=linspace(-0.5,0.5,250); n=2:2:106; i get the same error !
Please correct it for me. And show me where are the mistakes
0 个评论
回答(3 个)
Andrei Bobrov
2012-9-18
编辑:Andrei Bobrov
2012-9-18
one way
%W1:
n = (1:2:401).';
t = linspace(-0.5,0.5,250);
f = 4/pi*sum(bsxfun(@times,1./n,sin(2*pi*n*t)));
%W2:
n = (2:2:106).';
tau = linspace(-0.5,0.5,250);
f = 1/pi+bsxfun(@minus,0.5*sin(pi*tau),2/pi*(bsxfun(@rdivide,cos(pi*n*tau),n.^2)-1));
2 个评论
Jan
2012-9-18
Mohammed has written [MOVED from answer to comment section]:
But i need 250 variables for t
Jan
2012-9-18
for the expression n .* t the variables n and t must have the same size. Perhaps you want a matrix like, e.g. created by a dyadic product:
n(:) * t
But as long as you do not specify this, n .* t is simply a bug, when the sizes differ or non of them is a scalar.
0 个评论
Mohammed
2012-9-18
1 个评论
Jan
2012-9-18
编辑:Jan
2012-9-18
Please, Mohammed, do not post comments to answers as answer, but as comment. When you claim, that a suggested method "does not work", be so kind to show the code, the error message or an explanation of the difference between your expectation and the results.
Of course my suggestion will "not work", because I told you, that there is a bug in your code or that you at least did not tell us, what you want to achieve exactly enough. How could this be "working"?!
另请参阅
类别
在 Help Center 和 File Exchange 中查找有关 Matrix Indexing 的更多信息
Community Treasure Hunt
Find the treasures in MATLAB Central and discover how the community can help you!
Start Hunting!