ENH: cox proportional hazard, PHReg, survival function

[josef-pkt]

https://stackoverflow.com/questions/74403023/how-to-plot-survival-curve-for-cox-survival-from-statsmodels-api

guessing, I didn't look at it
We need to stitch it together, cox ph maximizes only partial likelihood

Hazard function should be just multiplication by proportionality factor.
But we want also survival function, How do we recover it from the hazard function?

I guess inference like conf_int for it will be very difficult combining baseline uncertainty with PH parameter uncertainty.