Using the findchangepts function on data containing NaN datapoints

Hi,

My data contains NaN datapoints, so when I use the findchangepts function, I get the following error message:

xError using findchangepts

Expected input number 1, X, to be finite.

Error in findchangepts>getargs (line 166)

validateattributes(x,{'single','double'},{'real','2d','nonsparse','finite'}, ...

Error in findchangepts (line 99)

[statistic, Kmax, Lmin, penalty] = getargs(x, varargin);

Is there a way to use the findchangepts function on data containing NaN datapoints?

If not, is there someway for me to transform my data so the findchangepts function can accept it without any errors?

Thanking you in advance!

Answer 1

[Samatha Aleti]

You may remove NaN values from your "data" and then call “findchangepts” function with this "data". You can remove NaN values from your “data” as follows:

data = data(~isnan(data)) 

[Impala]

Hi,

Thank you for your help.

If I remove the NaN values, the length of my data will reduce. I need to keep the length the same. Is there another way around this?

Many thanks once again.

Answer 2

[tara es]

you can use fillgaps function for fill NAN values (GAPs), without reduce the length of data. It can help you.

y = fillgaps(data)