Medium2

Import library and Data

I have used Lung dataset with sex being the only predictor.

library(xgboost)
library("survival")
data("lung")
lung[,"y"] = ifelse(lung[,"status"]==1,-lung[,'time'],lung[,'time'])

Results

param <- list(  objective   = "survival:cox",
                eta         = 0.01,
                max.depth   = 2,
                subsample   = 0.5,
                nthread     = 2
)


df_train  <- xgb.DMatrix(as.matrix(lung[,"sex"]), label = as.matrix(lung[,"y"]))
bstSparse <- xgb.cv(data = df_train, nrounds = 10, params = param, nfold = 5, showsd = F,prediction = T)

## [1]  train-cox-nloglik:4.325408  test-cox-nloglik:2.980906 
## [2]  train-cox-nloglik:4.324707  test-cox-nloglik:2.980228 
## [3]  train-cox-nloglik:4.324192  test-cox-nloglik:2.979716 
## [4]  train-cox-nloglik:4.323563  test-cox-nloglik:2.979122 
## [5]  train-cox-nloglik:4.322896  test-cox-nloglik:2.978465 
## [6]  train-cox-nloglik:4.322362  test-cox-nloglik:2.977881 
## [7]  train-cox-nloglik:4.321909  test-cox-nloglik:2.977393 
## [8]  train-cox-nloglik:4.321303  test-cox-nloglik:2.976775 
## [9]  train-cox-nloglik:4.320720  test-cox-nloglik:2.976190 
## [10] train-cox-nloglik:4.320322  test-cox-nloglik:2.975753

Sample output in console

Iter Number : 94 Prediction: -0.682183 True: 269
Iter Number : 95 Prediction: -0.682183 True: 270
Iter Number : 96 Prediction: -0.682183 True: -279
Iter Number : 97 Prediction: -0.682183 True: 284
Iter Number : 98 Prediction: -0.682183 True: 285
Iter Number : 99 Prediction: -0.703774 True: 285
Iter Number : 100 Prediction: -0.682183 True: 286
Iter Number : 101 Prediction: -0.682183 True: 288
Iter Number : 102 Prediction: -0.682183 True: 291
Iter Number : 103 Prediction: -0.703774 True: -292
Iter Number : 104 Prediction: -0.682183 True: -292