Question

Matlab拟合出错

clc
x1=[0.050.0750.10.20.40.60.20.20.20.20.20.2]';
x2=[54500054500054500054500054500054500054545.45181818.18363636.36909090.911454545.452181818.18]';
x=[x1,x2];
y=[769.9144325.0511174.300139.85486.9964032.46029142.524240.856139.854839.687839.64139.6194]';
fun=inline('1.445*1.225*x(:,1).^a(1)*x(:,2).^a(2)*(x(:,1)-1).^2','a','x');
a=lsqcurvefit(fun,[0 0],x,y)
提示错误信息为：
??? Error using ==> optim\private\lsqncommon
User supplied expression or inline function ==> 1.445*1.225*x(:,1).^a(1)*x(:,2).^a(2)*(x(:,1)-1).^2
failed with the following error:

Error using ==> inlineeval
Error in inline expression ==> 1.445*1.225*x(:,1).^a(1)*x(:,2).^a(2)*(x(:,1)-1).^2
??? Error using ==> mtimes
Inner matrix dimensions must agree.

Error in ==> lsqcurvefit at 149
[x,Resnorm,FVAL,EXITFLAG,OUTPUT,LAMBDA,JACOB] = ...

Error in ==> curvefitting at 7
a=lsqcurvefit(fun,[1 1],x,y)
>>
新手刚刚接触，查好久都没查到哪里的问题，跪求大虾出手看看哪里有问题，感激不尽!!!

Answer 1

出错原因，主要是在自定义函数fun（）上（没有很好理解点乘的用法）。正确的fun（）应为

fun=inline('1.445*1.225*x(:,1).^a(1).*x(:,2).^a(2).*(x(:,1)-1).^2','a','x');

根据你的数据，拟合结果为

Answer 2

把fun改成 fun=@(a,x)1.445*1.225*x(:,1).^a(1)*x(:,2).^a(2)*(x(:,1)-1).^2;
本来想给你运行出来的，你的x，y值没分开，连在一起了