Great, that's the best solution -- standardize on what you have in LogLikelihood. The other implementations should just use it. I have enough of a whiff of the intuition here to understand the other formulation and why it's a little more straightforward. I'll file a JIRA with patch and submit.
On Sun, Jan 30, 2011 at 8:45 PM, Ted Dunning <[email protected]> wrote: > > The result boils down to a few different forms. One form is the one I used > in my ancient paper (here: http://acl.ldc.upenn.edu/J/J93/J93-1003.pdf ). > That is similar to the form used in the code you have. Another form is the > one I pushed on my blog (here: > http://tdunning.blogspot.com/2008/03/surprise-and-coincidence.html ) > > I much prefer the second form because I can remember it precisely. It is > the form used in org.apache.mahout.math.stats.LogLikelihood > > Does that help?
