问我用这个msm包的代码拟合模型，结果总是显示失败，试了几种方法，都还是不能给出结果，我可不可以认为，这个变量不太适合这个模型？

提问于 2025-02-26 17:11:47

回答 1关注 0查看 47

cov1 <- c("NOx","ethnic","BMI")
> msm1<-msm(state ~ years, subject = f.eid, data = data8,qmatrix= matrixQ, death=4,
+           obstype= 1,gen.inits = TRUE,method ="BFGS",control=list(fnscale=500,maxit=40000),
+           covariates = as.formula(paste("~",paste0(cov1,collapse ="+"),sep="")))
警告信息:
In msm(state ~ years, subject = f.eid, data = data8, qmatrix = matrixQ,  :
  Optimisation has probably not converged to the maximum likelihood - Hessian is not positive definite.

msmq