' convert blrm objects to (Frequentist) orm objects (issue warning that poma is not yet implemented for cppo models)
' provides option to generate transformed 1-ECDF curves for model predictors (specified by the predvar argument) (This helps user decide if Bayesian cppo should be used)
' poma compares logit ECDF and inverse normal ECDF to show that ordinal regression is often a better option than linear regression (even when the PO assumptions are not fully met)
Dear Sir,
The updated poma.R performs the following:
' convert
blrmobjects to (Frequentist)ormobjects (issue warning that poma is not yet implemented for cppo models)' provides option to generate transformed 1-ECDF curves for model predictors (specified by the
predvarargument) (This helps user decide if Bayesian cppo should be used)' poma compares logit ECDF and inverse normal ECDF to show that ordinal regression is often a better option than linear regression (even when the PO assumptions are not fully met)
Thank you for your kind consideration.