Linear Regression with NaN Elements

Hi everbody,

I have a data set which is named as AA.

AA=[0.02 NaN NaN NaN 0.03 0.035 0.045 NaN NaN 0.05 0.059]

I should create a linear regression on NaN elements. (regression must be created from 0.02 to 0.03 and from 0.045 to 0.05 )

How can i done it?

Thanks.

[Birdman]

With NaN elements or without?

[Computer Lover]

New dataset mustn't include NaN elements. Regression values should replace NaN elements' locations.

Answer 1

[Birdman]

Something like this?(replacing NaN values with the mean of array without including NaNs)

AA(isnan(AA))=mean(AA(~isnan(AA)))