Codes for Mplus using MLE based on block-Toeplitz to estimate Dynamic factor models
2006-11-17    Zhang, Z.   
Print from: Zhiyong Zhang \'s Psychometric Website
Address: https://www.psychstat.org/us/article.php/71
Codes for Mplus using MLE based on block-Toeplitz to estimate Dynamic factor models

TITLE:        Simulation study DFMs!The input covariance matrix are calculated pooled lagged covariance matrix
!The fit statistic need to modified according to the degree of freedom

DATA:         FILE = single.txt;
              TYPE IS covariance;
              NOBSERVATIONS=99;

VARIABLE:     NAMES = y11 y21 y31 y41 y51 y61
                        y12 y22 y32 y42 y52 y62;
              USEVARIABLES = y11 y21 y31 y41 y51 y61
                        y12 y22 y32 y42 y52 y62;
                  

ANALYSIS:      estimator=ML;
MODEL:         !Lag 0
               f11 by y11*.5(1); f11 by  y21(2);  f11 by  y31(3);
               f21 by y41*.7(4); f21 by  y51(5);  f21 by  y61(6);
               !Lag 1
               f12 by y12*.5(1); f12 by  y22(2);  f12 by  y32(3);
               f22 by y42*.7(4); f22 by  y52(5);  f22 by  y62(6);
               !Equal residual variances
               y11(8); y21(9); y31(10); y41(11); y51(12);y61(13);
               y12(8); y22(9); y32(10); y42(11); y52(12);y62(13);

               f11 WITH f21;
               f12 WITH f22;
               !Autoregressive parts
               f12 on f11 (b1);
               f12 on f21 (b2);

               f22 on f11 (b3);
               f22 on f21 (b4);
               !factor variance and shock variances
               f12@.36;
               f22@.36;
               f11(p1);
               f21(p2);
OUTPUT:      RESIDUAL SAMP STAND TECH1;
SAVEDATA: Results are results.txt;

Editor: johnny