telensky wrote:Is there MARK source code available?
No...
(since there is no documentation
Really, no documentation? Relative to what?
and some features are really not documented well in the MARK book
A fair comment. Its a never-ending enterprise.
- for example, what is MARK doing with the dots in the capture history in the CJS models, how exactly is that handled statistically etc., I wanna see if this could be used for model validation which currently doesn't allow it etc... so basically I would like to see MARK source to understand more what it does.)
Unless you can read FORTRAN 90 (which is the language MARK is written in) at a very high level, the source code (all 46,833 plus lines of it, exclusive of the LINPACK libraries that are compiled in) isn't likely to reveal much that would help. And even if you do read FORTRAN (meaning, you're either >50 years old and come from an era where FORTRAN was the primary language, or a physicist, where FORTRAN is still the go-to language), 46,833 lines of code is a fair bit to 'read'.
If you really want to know how something works, simulate, experiment, and go from there. For CJS models, the 'hint' is the relationship between a 'missing interior samples' and 'unequal intervals' (which is covered in the book).