Re: Curve Fitting to timeseries using a set of 8 sine and cosine functions [message #89630 is a reply to message #89615] |
Sun, 02 November 2014 20:49   |
Craig Markwardt
Messages: 1869 Registered: November 1996
|
Senior Member |
|
|
On Friday, October 31, 2014 11:12:03 PM UTC-4, siumt...@gmail.com wrote:
> My question is not resolved yet.
>
> Craig: Why you did not suggest me to use your MPFIT code for fitting my model.
You can use MPFIT if you want. As you mentioned in the original post, you knew about MPFIT already, so why did I need to suggest it again?
In any case, since an FFT does produce a least squares set of fourier coefficients, using MPFIT is kind of overkill.
You said this,
> I have attempted to used multiple linear regression to solve the problem .
> However, when I plot the original data with fitted line, I did not find good
> result.
So - given your model - how do you know there is a better result?
Craig
|
|
|