I had a chat with Danny this morning about how to handle sim studies that have convergence issues and edited functions and vignettes according to what he said. Now, the calc_* outputs K which might differ by condition if there are NA values in place of performance estimates. The functions delete NA values before calculating anything. I have also gotten qs about convergence issues from Man and Kejin.
I had a chat with Danny this morning about how to handle sim studies that have convergence issues and edited functions and vignettes according to what he said. Now, the
calc_*outputs K which might differ by condition if there areNAvalues in place of performance estimates. The functions delete NA values before calculating anything. I have also gotten qs about convergence issues from Man and Kejin.I added the following to the Simulation Workflow vignette: https://meghapsimatrix.github.io/simhelpers/articles/simulation_workflow.html#convergence-issues
I also added this para under Absolute Criteria section in the MCSE vignette:
And, added this to the Relative Criteria for Variance estimators section: