Panel-Generator for Visualization of A Precision Medicine Tree
Panel-Generator for Visualization of A Precision Medicine Tree
This function provides a new plot method for dipm
and spmtree. It visualizes stratified treatment groups through boxplots for a continuous outcome and survival plots for a survival outcome, respectively.
node_dipm(obj,...)
Arguments
obj: A party tree object returned from either the dipm() or spmtree() function
...: Arguments passed on to plotfun
Returns
No return value, called for plot
Details
This function visualizes the precision medicine trees proposed in Chen and Zhang (2020a, b).
Examples
#' ## ... an example with a continuous outcome variable# and two treatment groups#N =100set.seed(123)# generate binary treatmentstreatment = rbinom(N,1,0.5)# generate candidate split variablesX1 = rnorm(n = N, mean =0, sd =1)X2 = rnorm(n = N, mean =0, sd =1)X3 = rnorm(n = N, mean =0, sd =1)X4 = rnorm(n = N, mean =0, sd =1)X5 = rnorm(n = N, mean =0, sd =1)X = cbind(X1, X2, X3, X4, X5)colnames(X)= paste0("X",1:5)# generate continuous outcome variablecalculateLink =function(X, treatment){((X[,1]<=0)&(X[,2]<=0))*(25*(1- treatment)+8* treatment)+((X[,1]<=0)&(X[,2]>0))*(18*(1- treatment)+20* treatment)+((X[,1]>0)&(X[,3]<=0))*(20*(1- treatment)+18* treatment)+((X[,1]>0)&(X[,3]>0))*(8*(1- treatment)+25* treatment)}Link = calculateLink(X, treatment)Y = rnorm(N, mean = Link, sd =1)# combine variables in a data framedata = data.frame(X, Y, treatment)# fit a dipm classification treetree = dipm(Y ~ treatment | ., data, mtry =1, maxdepth =3)plot(tree, terminal_panel = node_dipm)
References
Chen, V., Li, C., and Zhang, H. (2022). dipm: an R package implementing the Depth Importance in Precision Medicine (DIPM) tree and Forest-based method. Bioinformatics Advances, 2 (1), vbac041.
Chen, V. and Zhang, H. (2020). Depth importance in precision medicine (DIPM): a tree and forest based method. In Contemporary Experimental Design, Multivariate Analysis and Data Mining, 243-259.
Chen, V. and Zhang, H. (2022). Depth importance in precision medicine (DIPM): A tree-and forest-based method for right-censored survival outcomes. Biostatistics 23 (1), 157-172.
Seibold, H., Zeileis, A., and Hothorn, T. (2019). model4you: An R package for personalised treatment effect estimation. Journal of Open Research Software 7 (1).
Hothorn, T. and Zeileis, A. (2015). partykit: a modular toolkit for recursive partytioning in R. The Journal of Machine Learning Research