The goal: simulate chi square mixture distributions as a way of simulating likelihood ratio test statistics for some mixed models where the more specific model has some zero variance components (a la Pinheiro and Bates pg. 84-87) The problem: R doesn't have the function ms which is apparently used by simulate.lme In the current version of nlme for R, is there a way around this? Is it possible to use the simulate function? Is this going to be rectified in the (eagerly anticipated) new release of nlme next month? Thanks, Scott.rifkin at yale.edu
Scott Rifkin <scott.rifkin at yale.edu> writes:> The goal: simulate chi square mixture distributions as a way of > simulating likelihood ratio test statistics for some mixed models where > the more specific model has some zero variance components (a la Pinheiro > and Bates pg. 84-87) > > The problem: R doesn't have the function ms which is apparently used by > simulate.lme > > In the current version of nlme for R, is there a way around this? Is it > possible to use the simulate function? Is this going to be rectified in > the (eagerly anticipated) new release of nlme next month?You may be able to replace the call to ms with an appropriate call to optim or to nlm. I'd prefer not to take the time right now to work out exactly how that would be done. I have mentioned several times on this list that I'm in the process of developing a new and wonderful implementation of lme and I would prefer to continue working on that rather than modifying old-style code. If you want to try to make the modifications yourself I can respond to questions. -- Douglas Bates bates at stat.wisc.edu Statistics Department 608/262-2598 University of Wisconsin - Madison http://www.stat.wisc.edu/~bates/