plot.ensemble {randomSurvivalForest} | R Documentation |
Plot ensemble survival curves and ensemble estimates of mortality.
plot.ensemble(x, plots.one.page = TRUE, ...)
x |
An object of class (rsf, grow) or (rsf,
predict) . |
plots.one.page |
Logical. Should plots be placed on one page? Default is TRUE. |
... |
Further arguments passed to or from other methods. |
Four plots are produced. Going from top to bottom, left to right: (1) Ensemble survival function for each individual. Thick red line is overall ensemble survival, thick green line is Nelson-Aalen estimator. (2) Comparison of the population ensemble survival function to the Nelson-Aalen estimator. (3) Brier score (0=perfect, 1=poor, and 0.25=guessing) stratified by ensemble mortality. Based on the method described in Gerds et al. (2008) in which the censoring distribution is estimated using the Kaplan-Meier estimator. Stratification is into 4 groups corresponding to the 0-25, 25-50, 50-75 and 75-100 percentile values of mortality. Red line is the overall (non-stratified) Brier score. (4) Plot of mortality versus observed time. Points in blue correspond to events, black points are censored observations.
Note that when x
is of class (rsf, predict)
not all
plots will be produced.
For competing risk analyses plots correspond to unconditional values
(i.e., they are non-event specific). Use competing.risk
for
event-specific curves and for a more comprehensive analysis in such
cases.
Whenever possible, out-of-bag (OOB) values are used.
Hemant Ishwaran hemant.ishwaran@gmail.com and Udaya B. Kogalur ubk2101@columbia.edu
T.A. Gerds, T. Cai and M. Schumacher (2008). The performance of risk prediction models, Biometrical Jour., 4:457-479.
E. Graf, C. Schmoor, W. Sauerbrei and M. Schumacher M (1999). Assessment and comparison of prognostic classification schemes for survival data, Statistics in Medicine, 18:2529-2545.
H. Ishwaran, U.B. Kogalur (2007). Random survival forests for R, Rnews, 7/2:25-31.
competing.risk
,
rsf
,
predict.rsf
.
data(veteran, package = "randomSurvivalForest") v.out <- rsf(Survrsf(time, status)~., veteran, ntree = 1000) plot.ensemble(v.out) # plot of ensemble survival for a single individual surv.ensb <- t(exp(-v.out$oob.ensemble)) plot(v.out$timeInterest, surv.ensb[, 1])