Issues with death_rate_response.c

[nichole-carlson]

Same issue with death_rate_trigger and the use of parms_cox in this subroutine.

We also have use of full likelihood. This might imply likelihood for the whole population of subjects where it is only the likelihood for the single subject (I think---we should check).

[nichole-carlson]

Fix *.h file

[nichole-carlson]

I do trust that the calculation here is correct. I will have to verify the theory.