Linear Least Square fit

Hi,

I am new here and I apologize for any mistakes.

I have a problem and I don't know how to solve it please guide me.it"s not a homework or anything.

a data is given and

and y values have variance-covariance

find the optimal parameters c1 and c2 for fitting function

calculate the variance-covariance matrix.

[John D'Errico]

This is a bit confusing.

Are you asking to compute a least squares fit, GIVEN a known covariance matrix on the vector y? If so, then you can use lscov.

Or are you asking to compute a least squares fit, and then estimate a covariance matrix for the parameters?

First, you showed a covariance matrix, but then you ask to compute a covariance matrix.

[dariyoush shiri]

sorry for the confusion, I guess to compute a least squares fit, and then estimate a covariance matrix for the parameters. I got this question like I posted. and this is my first atempt on fitting of data so I am confused as well.

Answer 1

[Alan Stevens]

For the linear least squares best fit, try:

>> M = [ones(3,1), x.^3];
>> C = M\y

C =

    0.6667
    0.5000

For the var-covar look up the MATLAB documentation on cov.

[dariyoush shiri]

thank you for the answer, can you please tell me how can I calculate var-cov matrix it manually?