WPC (version 1.0)

DuoWPCCurve: Generate Two Weighted Predictiveness Curves in Graph

Description

This function will generate two weighted predictiveness curves using the estimates provided by "npr.wpc.est" or "cox.wpc.est" functions. It can be used to compare the relationships between survival rate and biomarker from two different curves.

We can utilize this function to compare the performance between non-parametric predictiveness curve and parametric(cox) predictiveness curve, or compare the performance from non-parametric predictiveness curves using two different sets of parameters, or compare the predictiveness curves by using data from two different treatment groups and therefore compare treatment-by-biomarker relationships.

Usage

DuoWPCCurve(wpc1, wpc2, xlab, ylab, main, ylim, xlim, type, col1, col2, lwd, 
legendloc, legendtxt, confi, ptsest)

Arguments

wpc1

It is the object1 generated by function cox.wpc.est or npr.wpc.est.

wpc2

It is the object2 generated by function cox.wpc.est or npr.wpc.est.

xlab

It is the title for x axis; default is "Marker".

ylab

It is the title for y axis; default is "Survival Rate".

main

It is the title for the plot; default is "Weighted Predictiveness Curve".

ylim

It creates the continuous scale of y axis of the plot; default is "c(0,1)".

xlim

It creates the continuous scale of y axis of the plot; default is "c(0,100)".

type

It defines the type of the curves; default is "l".

col1

It defines the color of the curve 1 from object 1; default is "red".

col2

It defines the color of the curve 2 from object 2; default is "blue".

lwd

It defines the width of the curve; default is "2".

legendloc

It specifies the location of the legend; default is "bottomright".

legendtxt

It provides the text of the legend; default is "c("Method1")".

confi

It provides the option of drawing the confidence bands; default is "N", which means no confidence band is needed; "Y" will report the confidence band.

ptsest

It provides the option of drawing the point estimates; default is "N", which means no point estimates is needed; "Y" will report the point estimates.

References

Yang H., Tang R., Hale M. and Huang J. (2016) A visualization method measuring the performance of biomarkers for guiding treatment decisions Pharmaceutical Statistics, 15(2), 1539-1612

See Also

SoloWPCCurve, TrioWPCCurve

Examples

	# Get the estiamte of predictiveness curve from npr.wpc.est functions 
	
	npr.object = npr.wpc.est(event=wpcdata$OSday, censor=wpcdata$OScensor, 
	marker=wpcdata$Biomarker1,cutoff=180,method="number.subjt",weights="normal",
	nsub=10,sspeed=1,df=2,confi="NO")
	
	# Get the estiamte of predictiveness curve from cox.wpc.est functions 

	cox.object = cox.wpc.est(event=wpcdata$OSday, censor=wpcdata$OScensor, 
	marker=wpcdata$Biomarker1,cutoff=180,quantile=0.95)

	# Print Predictiveness Curve

	DuoWPCCurve(npr.object,cox.object,xlab="Marker",ylab="Survival Rate",
	main="Weighted Predictiveness Curve",ylim=c(0,1),xlim=c(0,100),type="l",
	col1="red",col2="blue",lwd=2,legendloc="bottomright",
	legendtxt=c("treatment","placebo"),confi="N", ptsest="N")