Probability point estimate

[bachnguyen-tomo]

I want to give out a probability estimate of the event happening for each point in time, i.e P(t=x)=p, with x being any point in time. Since predict_survival_function() returns rows with values representing P(t > x), would it make sense to calculate the probabilty with P(t>x) - P(t>x-1)? Thanks in advance.