Using fit 'exp2' an error with "weights"

I've used the fit function a few times on creep data sets (i.e. time vs displacement) and have been using the 'exp2' fit type. On smaller data sets I've had no problem using this function. Now I'm looking at a 10 folder bigger data set and am getting an error.

I was previously running this and had no problem

[creep_data_fit, gof] = fit(creep_data(1:1800,1),creep_data(1:1800,2), 'exp2')

Now I'm running this and getting the following error

[creep_data_fit, gof] = fit(creep_data(1:10800,1),creep_data(1:10800,2), 'exp2')

Any ideas on why i'm getting this error?

Error using fit>iFit (line 235)
X, Y and WEIGHTS cannot have NaN values.

Error in fit (line 109)
[fitobj, goodness, output, convmsg] = iFit( xdatain, ydatain, fittypeobj, ...

Answer 1

[Tom Lane]

Since you didn't specify weights, it seems like the NaN value must be in either x or y. You can find the usable values (using your own variable names) like this:

t = ~isnan(x) & ~isnan(y);

Then omit the NaN values from the fit:

fit(x(t),y(t),'poly1')

[Alison]

This worked great. Thanks Tom.

Answer 2

[Alison]

This worked great thanks Tom!

[Zahra Ali]

I had been working on something similar for a very long time, thank you for your help!