Matrix Dimensions must agree using integral

Question

atsprink 2018-2-21

0
链接

此问题的直接链接

https://ww2.mathworks.cn/matlabcentral/answers/383872-matrix-dimensions-must-agree-using-integral

编辑： Torsten 2023-2-23

采纳的回答： atsprink

在 MATLAB Online 中打开

Hello all,

I am trying to integrate over a function.

% calculation of f(b)
b = exp((-q*Vj)/(2*kB*Temp))*cosh((Et-Ei)/((kB*Temp)+0.5*(log(tau_p0/tau_n0))));
% integrate over x while varying b
fcn_b = zeros(1,length(b));
fcn = @(x)1./((x.^2)+(2.*b.*x)+1);
fcn_b = integral(fcn,0,Inf);
end

But I receive an error after the fcn = @x line saying matrix dimensions must agree.

The size of b => 1 x 141. Any help would be appreciated.

2 个评论
显示无隐藏无

KSSV 2018-2-21

编辑：KSSV 2018-2-21

Give us the complete code......what is q? what is size of x?

atsprink 2018-2-21

编辑：atsprink 2018-2-21

All of those values are just constants except Vj. Vj is an array of 1 x 141. x is not a size, its a function of x as in f(x) = Integration of 1 / (x^2 + 2bx + 1) I uploaded all the files. Start with main.m

edit: added too many files.... comment out lines 10-19 in main.m as those function calls are not needed

请先登录，再进行评论。

请先登录，再回答此问题。

Answer 1

atsprink 2018-2-21

2
链接

此回答的直接链接

https://ww2.mathworks.cn/matlabcentral/answers/383872-matrix-dimensions-must-agree-using-integral#answer_306235

I figured it out I believe. I needed to add " 'ArrayValued', true " to my integral fcn_b line of code. It is at least not throwing an error right now but not sure if it is returning a correct answer.

3 个评论
显示 1更早的评论隐藏 1更早的评论

Lucas Ivan 2023-2-23

Hello everyone! I was having the same error and I could solve it with this. However, any idea on why we need to add this argument?

Torsten 2023-2-23

编辑：Torsten 2023-2-23

However, any idea on why we need to add this argument?

If you don't add this argument, MATLAB calls your function to be integrated with a vector of values for the integration variable and expects your function to return a vector of the same size. This won't happen since your function itself is a vector of functions - depending on the vector b. If you set "ArrayValued" to "true", MATLAB "knows" that you want to integrate a vector-valued function and calls it only with single (scalar) values for the integration variable.

请先登录，再进行评论。