Vectorization of for loop

Question

Safwana Razak 2021-7-15

0
链接

此问题的直接链接

https://ww2.mathworks.cn/matlabcentral/answers/879173-vectorization-of-for-loop

回答： Jayant Gangwar 2021-7-15

B = randi(10x5); 
x = randi(10x40); 
y = randi(10x1);
% Modelfun = equation to fit
% I can do a for loop like this:
for i=1:10
    [beta(i,:)]=nlinfit(x(i,:),y(i,:),modelfun,B)
end

Can I do vectorization? for example 10 fittings all at once, without using loops? or maybe using @cellfun or @arrayfun?

1 个评论
显示 -1更早的评论隐藏 -1更早的评论

KSSV 2021-7-15

编辑：KSSV 2021-7-15

cellfun, arrayfun also uses loop inside...

请先登录，再进行评论。

请先登录，再回答此问题。

Answer 1

Jayant Gangwar 2021-7-15

0
链接

此回答的直接链接

https://ww2.mathworks.cn/matlabcentral/answers/879173-vectorization-of-for-loop#answer_747308

在 MATLAB Online 中打开

It seems to me that you want to avoid the use of loops for finding all the rows of beta, You can do it by directly passing the complete x matrix and y vector to the nlinfit function, It will automatically save the answer in different rows of beta. An example of the same is given below-

S = load('reaction');
X = S.reactants; % 13x3 matrix
y = S.rate; % 13x1 vector
beta0 = S.beta;
[beta,R,J,CovB,MSE,ErrorModelInfo] = nlinfit(X,y,@hougen,beta0,'ErrorModel','combined');
beta

This is an example given in the documentation for nlinfit, for more information please take a look at the documentation for nlinfit - Nonlinear regression - MATLAB nlinfit (mathworks.com)