Function to fit survival data

James Barnes

2016 3 25

0 个回答

8 次查看（30 天）

0 个投票

I have two sets of survival data, for example: A = [1;.92;.88;.85;.80;.75] and B = [1;.95;.94;.93;.91;.89] corresponding to times (months): T = [0;1;2;3;4;5]. I am looking to fit a survival function to this data to get a Hazard ratios for B/A. It appears that coxphfit, mle, expfit for instance, require raw individual survival time data and frequency data as inputs, rather than the actual survival data. I haven't found exactly what I am looking for yet...

I guess what I currently have is a survival function (cdf) that i want to find some set of parameters for an exponential, weibull, or other distribution that fits this data, so i can derive the implied HR and find the point estimate for the ratios. I cannot easily get the raw data for this. Many different distributions or methods may work for me.

Thank you in advance!