[R] random truncation

[Rolf Turner]

On 13/07/19 3:31 PM, Abby Spurdle wrote:

>  > The distribution of the randomly truncated variable has thus four
>  > parameters: a, b, mu and sigma.  I was able to write down the likelihood
>  > and attempted to maximise it
> 
> I read the Wikipedia article more carefully.
> The formula is relatively simple, and is based on the application of 
> Bayes Theorem.
> If one doesn't want to work out the integral, numerical methods can be used.

I am mystified as to what you are saying here.  What integral?
As I said, I could write down the likelihood.  Explicitly.  Maximising 
it (numerically *of course*) turned out to be problematic.  Back then.
Modern optimisers might help.  Presumably the DTDA package has a 
reasonable procedure.  (I haven't looked at it.)

> However, the problem needs to be defined *precisely* first.

Again I have no idea of what you are driving at here.  The concept of 
"random truncation" is quite precisely defined.

cheers,

Rolf

-- 
Honorary Research Fellow
Department of Statistics
University of Auckland
Phone: +64-9-373-7599 ext. 88276