computes confusion matrix for (survival.time,censoring) outcome based on fit object "fit" and class predictions "yhat" soft response probabilities for (survival.time,censoring) are first estimated using Kaplan-Meier method applied to training data
pamr.confusion.survival(fit, survival.time, censoring.status, yhat)
The result of a call to pamr.train or pamr.cv
Survival time
censoring status
class predictions
Trevor Hastie, Robert Tibshirani, Balasubramanian Narasimhan, and Gilbert Chu