Hello, my question is concerning commands that could be used to make out of sample predictions.
Specifically, I am wondering if there are official stata commands to achieve the same or equivalent commands that are achieved using gllapred and ci_marg_mu lower upper, level(95) dots? This comes from a presentation given in 2008 by Sophia Rabe-Hasketh which I have attached and we see on page 21 what I am trying to achieve using Stata's mixed command and whatever post estimation commands that would match this. (my data are not binary but the same principles can be applied with proper specifications)...
The problem is basically that I have data that I want to predict based on a model and I want the ci around that prediction to include uncertainty because they are out of sample predictions etc. If there are other articles or textbooks that anybody here recommends please refer me to those.
I know it may be vague but the pdf is quite clear and very helpful.
thank you very much,
Jean-Michel Galarneau
Related Posts with Future/out of sample predictions
how to use firm new names to replace old names in two different dataset? Code: * Example generated by -dataex-. To install: ssc install dataex clear input str65 firmNameNo…
Generalized difference in differenceI'm studying the causal effects of a state policy, nurse practitioner scope of practice law, on vari…
Diagnostics: Power of ovtest and leverage pointsDear all, I have two relatively small Data sets (N=76 and N=111) and I use a WLS approach with awei…
Why do Poisson and Negative Binomial Regressions yield the same result?I am using the count data models as I have a count dependent variable (0, 1, 2, 3, 4). There is no …
Merge data set WhartonHello, I have downloaded two datasets from Wharton. One CRSP/Compustat Merged - Fundamentals Annual…
Subscribe to:
Post Comments (Atom)
0 Response to Future/out of sample predictions
Post a Comment