|
|
Missing convergence command in Versio... |
|
Message/Author |
|
|
Hi - I'm fitting a complex mixture model where I need to be able to control, for the initial runs, the convergence criteria. I've been using mconvergence and logcriterion. It says in the manual that logcriterion "is used to specify the absolute and relative observed-data loglikelihood change..." but it appears to actually only control the absolute criterion. I took a guess and tried using the command "rlogcriterion" and that works! I can't find it in the manual anywhere, is this a "missing" command? |
|
|
It's an undocumented command. |
|
Back to top |
|
|