05-23-2014 07:57 AM
Hi, I am looking to examine relationships between illness and contamination data. Given the data sparseness or zero counts in the contamination datatset, exploratory look using transfer functions did not quite prove what I was looking for so I am now looking into Brownian distance covariance for general correlation given the non-linear situation to assess dependence. I have searched the web and continuing to do so but has anyone identified a SAS procedure to conduct Brownian distance covariance? Any assistance would be greatly appreciated. Thank you
05-23-2014 10:22 AM
No current procedures. Looking at the MATLAB code, I suppose someone with a fair amount of time and bandwidth could convert it to PROC IML code, so you might try searching the web for something along those lines.