Phil Wood posted on Tuesday, November 04, 2008 - 7:32 pm
I'd like to request that Mplus have the ability to write files, much like the output files of parameter estimates, which specify which elements of which matrices are free and, if fixed, shows which elements are fixed to nonzero values. Something maybe like: phi(1,1) = p1 lambda(2,1)=1 etc. Just a thought- it would make some of the ancillary programming a lot easier- thanks for considering!
Are you asking to save TECH1? Or are you asking to save parameter estimates in the format of TECH1? I am not clear on your request.
Phil Wood posted on Sunday, November 09, 2008 - 1:30 am
Saving tech1 would work, or, alternatively, being able to just save something like the matrix name, element row and column, and parameter number (for tech 1) and the matrix name, element row and column and numerical (start) value would be handy. thanks for responding- I don't know how practical it would be to do.
zhang, ying posted on Wednesday, February 04, 2009 - 5:46 pm
I used "tech 1" and "tech 3" as output to see the estimated covariance matrix of the paramter estimates. It seemed that the default decimal places of these covariances are 3. Is there any way to save more decimal places of these numbers (since these numbers are rounded)? or even save the estimated covariance matrix as a seperate file, like what LISREL does? Thanks!
Thank you, but I should have mentioned that this laptop belongs to the job and I do not have access to any directory except C:\user\tom. My question is how and where in the command file can I give a path for the output.out file to be saved. I am running this from SAS.
I don't know what happens when you are running from SAS. When you run Mplus from the Mplus editor, the output goes to the directory you are running from or the directory specified for the input file if one is specified.
I was searching on line for lectures on Latent Class Analysis and was able to find some of your lectures from UCLA, but the quality is so bad, it is almost impossible to watch, the small portion that I was able to watch was great. I was wondering if there are lectures with good quality for me to watch.