|
|
Cholesky decomposition with many vari... |
|
Message/Author |
|
|
Hi, I am trying to estimate a 24x24 covariance matrix subject to the requirement that it be positive definite, so I am estimating its Cholesky decomposition. However given that I have 24 variables it would be too cumbersome to specify a script like e.g. this one: Data: file is 'data1.dat'; Variable: names are x1-x3; usev are x1-x3; Analysis: iter=5000; Model: x1(v1); x2(v2); x3(v3); x1 with x2 (c12); x2 with x3 (c23); x1 with x3 (c13); model constraint: new(p11 p12 p13 p22 p23 p33); v1=p11^2; v2=p12^2+p22^2; v3=p13^2+p23^2+p33^2; c12=p11*p12; c13=p11*p13; c23=p13^2+p23^2+p33^2; Is there a simple way to estimate a Cholesky decomposition for a large number of variables? Thanks, Sanja |
|
|
Mplus does not have an option for Cholesky decomposition. |
|
Back to top |
|
|