Adapting Monte carlo from WebNote 15 PreviousNext
Mplus Discussion > Latent Variable Mixture Modeling >
Message/Author
 Michelle Colder Carras posted on Wednesday, December 04, 2013 - 7:36 am
I’m getting the error THE POPULATION COVARIANCE MATRIX THAT YOU GAVE AS INPUT IS NOT POSITIVE DEFINITE AS IT SHOULD BE. I want to use count and continuous indicators with specified starting values and class sizes for 3 classes. The means/variances I want to use are 0.39(0.12) for the continuous variable and 4.73(7.3), 4.36(18.23), 5.31(35.22)and 9.52(58.37) for the count variables. Could you please help me figure out what’s going on? Thanks.
~Michelle

Montecarlo:
Names are u1-u6 y;
generate u2-u6 (pi);
count=u2-u6 ;
Genclasses = c(3);
Classes = c(3);
Nobservations = 500;
Nrep = 1;
Auxiliary = y(DU3STEP);
Analysis: Type = Mixture;
Model Population:
%Overall%
[y@0];
y@1;
[c#1*0.5];
[c#2*0.3];
%c#1%
[u1*0.1 u2-u6*1];
[y*0];
%c#2%
[u1*0.38 u2-u6*4];
[y*5];

%c#3%
[u1*0.62 u2-u6*9];
[y*10];

Model:
%Overall%
[c#1*0.5];
[c#2*0.3];
[y] (1); y (2); ! This command is...

%c#1%
[u1*0.1 u2-u6*1];

%c#2%
[u1*0.38 u2-u6*4];

%c#3%
[u1*0.62 u2-u6*9];
 Linda K. Muthen posted on Wednesday, December 04, 2013 - 9:09 am
Please send the output and your license number to support@statmodel.com.
Back to top
Add Your Message Here
Post:
Username: Posting Information:
This is a private posting area. Only registered users and moderators may post messages here.
Password:
Options: Enable HTML code in message
Automatically activate URLs in message
Action: