How can I fit data into Gaussian model?

I have a matrix of 38-by-50 which represents 38 sample and 50 genes reading. I need to cluster this data matrix using Gaussian mixture model. Please, how can I fit this data with GMM by using gmdistribution? Thanks

Answer 1

[Shashank Prasanna]

That's a very small dataset. Do you know how many clusters are there? Did you try:

n = numbers_of_clusters
obj = gmdistribution.fit(X,n)

Use the examples from the documentation:

http://www.mathworks.com/help/stats/gmdistribution.fit.html

[Ahmed Alabdaly]

Shashank,many thanks for your answer but I think there is a misunderstanding to my question. In this example, we produce the data from the values of (mu,sigma) but in my case I have these data which is 38-by-50 matrix. I couldn't apply the function gmdistribution.fit(X,K) unless for k=1! I want to cluster this matrix by using Gaussian mixture model. Thanks

[Marc]

No you misunderstood I think... 38 experiments with 50 data points is a small data set. The stats toolbox will suffice just fine. Show some data and code where you are having some issues and maybe we can help.

[Ahmed Alabdaly]

Thanks, I've attached the data file. I've used the code obj=gmdistriution.fit(X,2) but there is error, in case of n=1 its ok but only one cluster can I get.