Does fscmrmr function stop after 49 features?

Hi all,

I'm trying to use the fscmrmr function to reduce the number of features from +600 down to about 180. I used the fscmrmr function and had a couple of problems.

• in either case, the indx returns only 49 features. After the 49th feature, it just lists the remaining features in numerical order. When looking at the scores, I see that after the 49th feature, the scores drop to 0.
• depending on if I use a table or if I use a matrix, I'm also getting different answers, which I find very weird. Why would the format matter?

[Gouri Chennuru]

Can you provide the data on which you have applied the fscmrmr function ?

Answer 1

[Aryan Safakish]

Here is the attached data. Thanks for your assistance

Answer 2

[Gouri Chennuru]

Hi Aryan,

The fscmrmr function computes the mutual information values for both categorical (discrete) and continuous variables. fscmrmr discretizes a continuous variable into 256 bins or the number of unique values in the variable if it is less than 256. The function finds optimal bivariate bins for each pair of variables using the adaptive algorithm

You can refer to these links in order to resolve the issue.

https://www.mathworks.com/help/stats/fscmrmr.html#mw_b2ff44e4-6244-46f1-9348-35ee0615b91b

https://www.mathworks.com/help/stats/fscmrmr.html#namevaluepairs