Hi, I´m estimating a Random survival forest with the package “randomForestSRC” and “ggrandomforests”. If I understand it correctly, the most sensible approach to evaluate predictions is by the “ concordance index” , AKA “ C - index ”. I have difficulties to understand how to get the metric. There are packages available, e.g. “Hmisc”, “pec”, “survcomp”, but all approaches seem rather complicated. Have you any suggestion to get it the easiest way?
When you run a Random Survival Forest, you do get an “error rate”, but I´m not sure what that is.