Counting Process Parametric Fitter

[RubyT336]

I am looking to implement a counting process model within the custom parametric model functionality provided by ParametricRegressionFitter, Looking at some previous issues, there has been some talk around recurrent events but nothing yet implemented. Is that indeed the case?

I had a think about creating a CountingProcessParametricRegressionFitter that could do this which inherits from ParametricRegressionFitter. As far as I could tell, I would need to introduce a way to receive data in a (tstart, tstop, event) format, then alter the cumulative hazard/likelihood such that I (i) loop through each subject (ii) loop through a series of time-intervals and (iii) add each time/subject iteration to the cumulative intensity/cumulative hazard function. Any thoughts on this method?

Fishing for any tips here - any feedback much appreciated!