Functionality for composite distributions

Open jtbuch opened this issue 3 years ago • 1 comments

I am trying to write up a TensorFlow distribution for the composite lognormal-Pareto model introduced here: https://www.tandfonline.com/doi/abs/10.1080/03461230601110447?journalCode=sact20

Is there a general recipe for writing up new distributions? In particular, I'm having trouble writing the sample() method.

Any help/advice will be appreciated!

Edit: There's a public R package for such distributions as documented here.

Mar 30 '22 13:03 jtbuch

What issue are you running into implementing sample? Can you maybe post some code that isn't working but you think should?

Mar 31 '22 16:03 csuter