Re: Curve Fitting and Plotting

[Fredrik Lingvall]

Ben Abbott wrote:
> On Sep 26, 2008, at 6:32 AM, Kearan Mc Pherson wrote:
>
>   
>> Hi
>>
>> I am new to Octave and Gnuplot. I need help on fitting a curve  
>> through a data.dat file, that contains values x an y in column format.
>> Any ideas where to start?
>>     
Can you describe your problem i little bit more? If a "quick-and-dirty"
method, like using polyval, is not sufficient then I can recommend
reading, for example, "Bayesian Interpolation", David J. C. MacKay,
Neural Computation, May 1992, Vol. 4, No. 3, Pages 415-447.

https://eprints.kfupm.edu.sa/28027/1/28027.pdf

/F

>
> If the data is in two columns, and you'd like a simple polynomial  
> fit ...
>
> ## Load the data into (x,y)
> data = load ('data.dat');
> xdata = data(:,1);
> ydata = data(:,2):
>
> ## Fit a 2nd order polynomial
> order = 2;
> p = polyfit (xdata, ydata, order);
>
> ## Evaluate the fitted polynomial
> x = linspace (min(xdata), max(xdata), 101)
> y = polyval (p, x);
>
> ## Plot
> plot (x, y, '-' xdata, ydata, 's')
> legend ('Fitted polynomial', 'Original Data')
>
> Ben
>