rcITR-package: Risk Controlled ITR Discovery

In kdoub5ha/rcITR: Risk Controlled ITR Discovery

Description

Tool for risk controlled ITR discovery

Details

The DESCRIPTION file: This package was not yet installed at build time.

Index: This package was not yet installed at build time.
Tool for discovery of risk controlled individualized treatment rules

Author(s)

Kevin Doubleday

Maintainer: Kevin Doubleday <[e-mail]>

References

[In Submission] Doubleday, K., Zhou, J., Fu, H. (2020), "Risk Controlled Decision Trees and Random Forests for Individualized Treatment Recommendations".

Examples

  ## Not run: 
    set.seed(123)
    # Generate example data
    dat <- generateData(n = 1000)
    
    # Construct rcDT
    tre1 <- rcDT(data = dat, 
                 split.var = 1:10,
                 risk.threshold = 2.75,
                 lambda = 1,
                 efficacy = "y",
                 risk = "r",
                 col.trt = "trt",
                 col.prtx = "prtx")
    
    # Prune and select rcDT
    pruned1 <- prune(tre1, a = 0, risk.threshold = 2.75, lambda = 1)
    rcDT.fit <- rcDT.select(dat = dat, 
                            split.var = 1:10,
                            lambda = 1,
                            risk.threshold = 2.75, 
                            efficacy = "y",
                            risk = "r",
                            col.trt = "trt",
                            col.prtx = "prtx",
                            nfolds = 5)
    
    # Construct rcRF
    set.seed(2)
    rcRF.fit <- rcRF(dat, 
                     split.var = 1:10, 
                     efficacy = "y", 
                     risk = "r",
                     col.trt = "trt",
                     col.prtx = "prtx",
                     risk.threshold = 2.75,
                     ntree = 100, 
                     lambda = 0.5)
    
    # Variable importances
    VI <- Variable.Importance.ITR(rcRF.fit, sort = FALSE)
    
    # Predictions
    preds.rcDT <- predict.ITR(rcDT.fit$best.tree.alpha, 
                              new.data = dat, 
                              split.var = 1:10)
    preds.rcRF <- predict.ITR(rcRF.fit, 
                              new.data = dat, 
                              split.var = 1:10)
  
## End(Not run)

kdoub5ha/rcITR documentation built on Aug. 5, 2020, 9:05 p.m.