Errors returned from skewness and kurtosis

I am simply calculating four moments of a time series. But I have problem using skewness and kurtosis.

Following this example on the web http://www.mathworks.co.uk/help/stats/skewness.html, I simply copied X = randn([5 4]) and y = skewness(X) on to command window. Matlab returns:

"Error using nanmean Too many input arguments.

Error in skewness (line 39) x0 = x - repmat(nanmean(x,dim), tile);"

Can anyone tell me what is going wrong? Thanks a lot.

I don't want to change any source code before I am sure what is wrong. I think 'nanmean' works just fine itself.

Answer 1

[Sean de Wolski]

It sounds like you have your own nanmean function shadowing the Statistics Toolbox' nanmean

which -all nanmean

Will show the offending file. Rename it to something else so that it does not conflict with Stats' nanmean.

[harshit Sapkal]

'nanmean' not found.

I got this after running the command what should I do

[Steven Lord]

If you don't have nanmean installed you probably don't have Statistics and Machine Learning Toolbox installed. You can check that using the ver function. If you don't have this toolbox installed you also don't have the skewness and kurtosis functions as they're part of this toolbox as well.

But if you just want to take the mean of an array omitting NaN values you can call the mean function in MATLAB with the 'omitnan' option.

Answer 2

[yvette-hyq]

Thanks. Yes. There is another nanmean function in Dynare, the program package I used.