I'm trying to understand the mathematics behind predict_survival_function(), predict_median(), predict_expectation(), and other functions related to prediction in CoxPHFitter. I'm able to apply these functions to my data to predict the survival function and an individual survival time. Nonetheless, I'm curious about how this magic takes place given that I've learned CoxPH Model is commonly used to predict hazard ratio rather than survival function or survival time (I got a superficial instinct that this has to be done by selecting specific distributions).
Could anyone help me with this? Any help is appreciated!
