Hazard specification of parametric baseline hazard

[seabbs]

At the moment the parametric baseline hazard is calculated by using the CDF of the distributions to get the discrete probability of report. It may be possible to instead directly specify this as a hazard. This may be optimal when combined with other hazard models (like proportional changes in hazard over time etc).

[seabbs]

Some things to think about:

https://becarioprecario.bitbucket.io/inla-gitbook/ch-survival.html

[seabbs]

I've looked into this a bit more and it definitely is very possible for continuous-time models. What I am not sure is possible is the discrete version but still looking.

[seabbs]

Done in #134