nonlinear curve fitting: how to add an extra data vector?

Hi, I know how to fit a curve to the ' ydata ' vector, e.g.:

opt = fitoptions('Method','NonlinearLeastSquares',...
               'Startpoint',[1,0.2]);
fun = fittype('p1*cos(x)+p2*sin(x)','options',opt);

[fitobject,gof,output] = fit(time',ydata,fun)

However, I would like to include an additional data vector [same size as time and ydata(time) ] as a non-free time varying parameter [ pdata(time) ] to the function, so that we have something along this form:

fun = fittype('pdata*(p1*cos(x)+p2*sin(x))','options',opt);

Does anyone know how to implement this?

Answer 1

[Matt J]

You can specify the model as an anonymous function and pass parameters to that in the usual way and/or you could use the 'problem' input option to fittype/fit, see

http://www.mathworks.com/help/curvefit/fittype.html#btpaend-18

[Ramirez]

I tried your suggestion using:

fun = fittype( @(p1,p2,pdata,x) pdata*(p1*cos(x)+p2*sin(x)))

[fitobject,gof,output] = fit(time',y,fun,'problem','pdata')

where pdata is a vector with the same length as y and time. However I get the following error message:

Error using fit>iAssertNumProblemParameters (line 1115) Wrong number of values for problem parameters. Specify the values as a cell array with one element for each problem parameter in the fittype.

[Matt J]

fun = fittype( @(p1,p2,x) pdata.*(p1.*cos(x)+p2.*sin(x)),'problem','pdata');

[fitobject,gof,output] = fit(time',y,fun,'problem',pdata)