I am trying to estimate cure model (long-term survivor model) using "strsmix" and "strsnmix" in Stata. It took me quite a while to realize that good starting values are crucial for quick convergence, something I only thought of when estimating aML or Mplus models.
No comments:
Post a Comment