如何为endorse模型计算各协变量的边际效应？

我正在为以下贝叶斯模型计算各协变量的边际效应——该模型采用MCMC算法，用于估算对民兵组织的支持度：

endorseFULL <- endorse(Y = Y, data = data_slvk, identical.lambda = FALSE, covariates = TRUE, formula.indiv = formula( ~ age + gender + education + is_capital + ideology + income + DemPolGrievence + PolicyPolGrievence + EUPolGrievence + EconGrievenceRetro + EconGrievenceProspInd + EconGrievenceProspAgg + NatPride + NativeRights + NativeJobs + DemonstrateNational + LawOrder + Chauvinism + ChristianSchool + DemonstrateTrad + GayNeighbor+ GayPartner+ ForNeighbor + ForPartner + Ukraine), hierarchical = FALSE)

涉及的协变量如下：

vars <- c("id", "gender", "age", "education", "is_capital", "income", "ideology", "DemPolGrievence","PolicyPolGrievence","EUPolGrievence","EconGrievenceRetro","EconGrievenceProspInd","EconGrievenceProspAgg","NatPride","NativeRights", 
"NativeJobs","LawOrder","Chauvinism","ChristianSchool","GayNeighbor","GayPartner","ForNeighbor","ForPartner","Ukraine", "DemonstrateNational", "DemonstrateTrad")

我已经尝试了几种方法，但都遇到了兼容性问题：

• 使用marginaleffects包计算，发现该包与endorse模型不兼容
• 尝试endorse包自带的predict函数，但它不支持identical.lambda = FALSE的可变lambda模型
• margins包同样与该模型类型不兼容

内容的提问来源于stack exchange，提问作者user25699393