Models fail to converge, despite large sample size and variables with only 2 categories

[mmanus]

Hi, I'm having a problem that seems to be happening to others as well. I'm running the following model on a dataset of 100 samples, and the categorical variable that I'm specifying only has 2 categories. So far I'm unable to make this work with bbdml either, but I'm not completely sure that I'm setting up that model correctly.

Code:

family <- differentialTest(formula = ~ mom_bmi_best_cat2, phi.formula = ~ mom_bmi_best_cat2, formula_null = ~ 1, phi.formula_null = ~ mom_bmi_best_cat2, test = "Wald", boot = TRUE, B=5, data = manitoba_corncob_family, fdr_cutoff = 0.05)

Error:

Error in differentialTest(formula = ~mom_bmi_best_cat2, phi.formula = ~mom_bmi_best_cat2, :
All models failed to converge!

I have also tried this code based on reading others' conversations:

family <- differentialTest(formula = ~ mom_bmi_best_cat2, phi.formula = ~ mom_bmi_best_cat2, formula_null = ~ 1, phi.formula_null = ~ mom_bmi_best_cat2, test = "Wald", boot = TRUE, B=5, data = manitoba_corncob_family, fdr_cutoff = 0.05, init=rep(1,4))

But I receive this error:

Error in differentialTest(formula = ~mom_bmi_best_cat2, phi.formula = ~mom_bmi_best_cat2, :
argument 10 matches multiple formal arguments

For what it's worth, I successfully ran corncob a few months ago on a different dataset with a much smaller sample size (N=30), and those models included multiple covariates. I'm not attaching my files here as they were shared by a collaborator, but I can reproduce this error with another dataset and share those files, if that is helpful. Thanks in advance for any insight!

• Melissa