Linear Inequality Constraint on param... PreviousNext
Mplus Discussion > Categorical Data Modeling >
Message/Author
 Sanjoy posted on Thursday, May 19, 2005 - 5:00 am
Dear Professor ... my intension is to put the restriction i.e. 4th coefficient is always greater than the first one and the second one is always greater than the third one and then do the ML

CATEGORICAL ARE Y1;

MODEL:
Y1 on
C23(P1)
C43(P2)
D32(P3)
D43(P4)
S B1 X1 X2 X3 X4;

MODEL CONSTRAINT:
P4 GE P1;
P2 GE P3;

I am getting this output once I run this
"
*** FATAL ERROR
Internal Error Code: PR1002. Error in the constraints model. An internal error has occurred. Please contact us about the error, providing both the input and data files if possible."

can you suggest me please how to deal with inequality constraint in MPlus

thanks and regards
 Linda K. Muthen posted on Thursday, May 19, 2005 - 12:40 pm
Please send your input/output, data, and license number to support@statmodel.com.
 Sanjoy posted on Thursday, May 19, 2005 - 7:07 pm
thanks madam .. i have just sent them
 Maksim Rudnev posted on Wednesday, April 02, 2014 - 10:24 pm
I am doing a multiple group LCA and would like to put inequality constraints on the differences of response probabilities across groups. The model is the following:

CLASSES = manifest(2) latent(2);
KNOWNCLASS = manifest(var1 = 1 2);
CATEGORICAL ARE var2-var4;

ANALYSIS:
TYPE = MIXTURE;
ESTIMATOR = MLR;
ALGORITHM=INTEGRATION;

MODEL:
%overall%
latent ON manifest;
%manifest#1.latent#1%
[var2$1-var4$3] (a1-a9);
%manifest#1.latent#2%
[var2$1-var4$3];
%manifest#2.latent#1%
[var2$1-var4$3] (b1-b9);
%manifest#2.latent#2%
[var2$1-var4$3];

MODEL CONSTRAINTS:
NEW (high low);
high = 0.05;
low = -0.05;
high > EXP(a1)/(1+EXP(a1)) - EXP(c1)/(1+EXP(c1));
low < EXP(a1)/(1+EXP(a1)) - EXP(c1)/(1+EXP(c1));


This code works smoothly, however when I increase the number of the same kind of inequality constraints I keep getting the following error message:
THE MODEL ESTIMATION DID NOT TERMINATE NORMALLY DUE TO AN ERROR IN THE
COMPUTATION. CHANGE YOUR MODEL AND/OR STARTING VALUES.

I realize that the model has become very complex, but is there any way to get rid of this error message?
 Bengt O. Muthen posted on Wednesday, April 02, 2014 - 11:16 pm
We need to see the output. Please send input, output, data and license number to Support.
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: