Possibly world's most boring picture, but nonetheless a start.
Today I posted the R source code for Log-logistic Mixture model on Google Code
http://code.google.com/p/loglogisticmixture/
This is one of the graphs it generates -- a simulated dataset of 500 data points from 3 loglogistic distributions. Estimated using the Maximum Likelihood method. Not too bad.

No comments:
Post a Comment