Message-ID: <E0079AD1-F7BB-44BE-B769-FD1D97E80947@web.de>
Date: 2015-05-07T08:22:57Z
From: Holger Steinmetz
Subject: Problem with using flexmix for regression mixtures
Hi there,
I would like to conduct a mixture regression analysis with the flexmix
Package. Was just playing around with the function stepFlexmix() and did not get a foot
into the door.
When I run the stepFlexmix-function, I get the following error (actually a
list of errors that repeats this sentence:
Error in FLXfit(model = model, concomitant = concomitant, control = control,
:
26 Log-likelihood: Inf
The X-Variable is very skewed (percentages of females in top management
teams; many zeros); missing data were omitted. The code was:
M1 <- stepFlexmix(rel_perf ~ prozfem, data = cdata2, k = 1:5, nrep = 5)
I would appreciate any hint what the problem might be.
Thanks in advance,
Holger