I am having trouble with the chains properly mixing in my model (modified from ZINB-WaVE in Stan for scRNA-seq analysis — What do you mean "heterogeneity"? ). In particular, the regression parameters for the bernoulli component of the zero-inflated negative binomial model (beta_pi) is really badly mi…

What are the values of the covariates? In particular, is the covariate associated with beta_pi[2] entirely positive? If so, then it looks to me like what’s happening is that the model sees no zero inflation as a plausible model configuration. Thus, the chains for beta_pi[2] are free to wander to …

Help with ZINB model

Modeling

malonzm1 February 12, 2022, 8:12am 7

Thanks!

Topic		Replies	Views
Predicting from a zero inflated negative binomial model Modeling specification	2	802	May 23, 2023
ZINB model convergence issues when predicting shape parameter brms	7	1055	January 26, 2020
Partially pooled beta binomial model Modeling	9	1807	April 12, 2019
Priors in a ZINB model Modeling fitting-issues , brms	0	39	March 17, 2025
Problems fitting a Multilevel Negative Binomial Model Modeling fitting-issues	1	651	November 6, 2020

Help with ZINB model

Related topics