TrioWPCCurve: Generate Three Weighted Predictiveness Curves in Graph

In WPC: Weighted Predictiveness Curve

Description

This function will generate three weighted predictiveness curves in graph 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 three different curves.

similarly, We can utilize this function to compare the performances from non-parametric predictiveness curves using three different sets of parameters, or compare the predictiveness curves by using data from three different treatment groups and therefore compare treatment-by-biomarker relationships.

Usage

TrioWPCCurve(wpc1, wpc2, wpc3, xlab, ylab, main, ylim, xlim, type, col1, col2, col3, 
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.
wpc3	It is the object3 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".
col3	It defines the color of the curve 3 from object 2; default is "black".
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.

Author(s)

Hui Yang [e-mail], Rui Tang [e-mail] and Jing Huang [e-mail]

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, DuoWPCCurve

Examples

	tmppb = wpcdata[wpcdata$TRTA=="Placebo",]
	tmptrt1 = wpcdata[wpcdata$TRTA=="LowDose",]
	tmptrt2 = wpcdata[wpcdata$TRTA=="HighDose",]
	o.data1 =data.frame(event=tmppb$OSday, censor=tmppb$OScensor, marker=tmppb$Biomarker1)
	o.data2 =data.frame(event=tmptrt1$OSday, censor=tmptrt1$OScensor, marker=tmptrt1$Biomarker1)
	o.data3 =data.frame(event=tmptrt2$OSday, censor=tmptrt2$OScensor, marker=tmptrt2$Biomarker1)
	
	## Not run: npr.object1 = npr.wpc.est(event=o.data1$event, censor=o.data1$censor, 
	marker=o.data1$marker,cutoff=180,method="window.width",weights="huber",
	wdth=10,sspeed=1, df=2, confi="YES", nbtsp=1000)
	npr.object2 = npr.wpc.est(event=o.data2$event, censor=o.data2$censor,
	marker=o.data2$marker,cutoff=180,method="window.width",weights="huber",
	wdth=10,sspeed=1, df=2, confi="YES", nbtsp=1000)
	npr.object3 = npr.wpc.est(event=o.data3$event, censor=o.data3$censor, 
	marker=o.data3$marker,cutoff=180,method="window.width",weights="huber",
	wdth=10,sspeed=1, df=2, confi="YES", nbtsp=1000)

	TrioWPCCurve(npr.object1,npr.object2,npr.object3,xlab="Marker",ylab="Survival Rate",
	main="Weighted Predictiveness Curve",ylim=c(0,1),xlim=c(0,100),type="l",col1="red",
	col2="blue",col3="black",lwd=2,legendloc="bottomright",legendtxt=c("Method1",
	"Method2","Method3"),confi="Y")
## End(Not run)

WPC documentation built on May 2, 2019, 6:52 a.m.