Handle row of missing continuous data in joint FIML objective

mspiegel's picture
Category:bug report

See http://openmx.psyc.virginia.edu/thread/1111


mhunter's picture


This still causes an error.

Running ACE
Error: The job for model 'ACE' exited abnormally with the error message: MxComputeGradientDescent: fitfunction ACE.fitfunction evaluated to 1.#INF00 ()
In addition: Warning message:
In model 'ACE' Optimizer returned a non-zero status code 6. The model does not satisfy the first-order optimality conditions to the required accuracy, and no improved point for the merit function could be found during the final linesearch (Mx status RED)

bivariate_ACE_w_fakedata.R 6.08 KB
fake_dzData.csv 2.28 KB
fake_mzData.csv 910 bytes
tbates's picture


Should be closed now?
The error was just not telling people the starting values were infeasible?

mhunter's picture


Perhaps a slight bug still. Bates is right that the starting values are infeasible. This is returned correctly now. However, with intervals=TRUE it currently returns the error

Error in runHelper(model, frontendStart, intervals, silent, suppressWarnings,  : 
  MxComputeConfidenceInterval: reference fit is not finite

because it tries to compute confidence intervals even though there's a status 10 and the base fit function value is INF/NaN.

jpritikin's picture


> However, with intervals=TRUE it currently returns the error

Should be fixed with 3e864c52

AdminHunter's picture


Status:active» closed