Mplusautomation when using bayes estimation does not save or extract the warning/error message if model does not converge

[gitz13]

I am running a simulation and noticed that one of iteration out of a 1000, was just blank. When I ran the the seed for that single data generation and analyzed the data in Mplus, I got an error " THE CONVERGENCE CRITERION IS NOT SATISFIED. INCREASE THE MAXIMUM NUMBER OF ITERATIONS OR INCREASE THE CONVERGENCE CRITERION. "

I have programmed all $errors and $warnings (there are 2 different warnings mplusautomation saves) to flag iterations but the package is not recognizing the elements in this error/warning that Mplus reports. So instead every parameter and summary estimate is blank with no errors/warnings

, see the below screen shot. A value of 1 for my warnings and errors means that they are blank. The reason this is happening is because I set the maximum number of Bayesian iterations to 100k, and this single iteration of my simulation required 109k Bayesian iterations. The other 999 simulations executed as expected.

Ultimately I just want to provide feedback to improve the mplusautomation package. I have attached the output file. When you specify readmodels(model)$errors and readmodels(model)$warnings you will see that the package is not recognizing the error.

I realize this is not a top priority issue but when using mplusautomation to run a simulation, users would expect that the package would recognize all the errors in the analysis.

884threeitemnoninformative.txt

[gitz13]

I have now run into mplusautomation not recognizing another error message when utilizing maximum likelihood estimation. Mplus is printing an error: THE MISSING DATA EM ALGORITHM FOR THE H1 MODEL HAS NOT CONVERGED WITH RESPECT TO THE PARAMETER ESTIMATES. THIS MAY BE DUE TO SPARSE DATA LEADING TO A SINGULAR COVARIANCE MATRIX ESTIMATE. INCREASE THE NUMBER OF H1 ITERATIONS.

 NOTE THAT THE NUMBER OF H1 PARAMETERS (MEANS, VARIANCES, AND
 COVARIANCES) IS GREATER THAN THE NUMBER OF OBSERVATIONS.
           NUMBER OF H1 PARAMETERS :        152
           NUMBER OF OBSERVATIONS  :         50

I am aware that it is an extremely small sample size and in practice should not be using maximum likelihood estimation. My problem is I am running a simulation and would like mplusautomation to recognize the error so that I can filter out this replication from my results. Instead I'm still getting parameter estimates and some fit statistics, but not all fit statistics.

Here is my text file if you want to see why the error message is not reading into mplusautomation. 121.issue.eightitemml.txt.out.txt

I already have a solution in R for my simulation for both issues but I realize that reading in all errors and warnings using mplusautomation would be most appropriate for the package. I am just sharing the issue as it might be an easy fix for the pacakge.