Vis_3d: Visulization of Probability, A-arm Selection Fraction and...
In ryamada22/SelfDecABP:

Description Usage Arguments Details Value Examples

Visulization probabilities against the A-arm fraction and Overall success rate plane via persp()

1	Vis_3d(g, title, zlims)

`g`	grid matrices outputed from the function dist.2D ()
`title`	the name of the strategy. E.st or T.st
`zlims`	the minimum and maximum z-axis.

x-axis indicates A-arm fraction and y-axis indicates the overall success rate, and z-axis indicates the probability.

perspective plot with surface

library(combinat)
N <- 50
out<-serialTables(N)
ABSF<-sapply(out$Z, unlist)  
 
A.cnt <- relist((ABSF[,1]+ABSF[,2]),out$sk)
A.fraction<-apply(ABSF,1,A.frac)
A.fraction[1] <- 0.5
Success.cnt <- relist(ABSF[,1]+ABSF[,3],out$sk)
Success.fraction <-success.fraction<-apply(ABSF,1,success.frac)
Success.fraction[1] <- 0.5

E.st_prob.A<-apply(ABSF,1,E.st_utinity)
#Probability to select the A-arm based on utinity function of E.st: E.st_utinity()
a<-0.8
#ture success rate of A-arm treatment
b<-0.6
#ture success rate of B-arm treatment
E.table.probability<-table.prob(relist(E.st_prob.A,out$sk),p=c(a,b))
#exact probability per 2X2 table based on utinity function of E.st: E.st_utinity()
grid.exp.ls<-dist.2D(Success.cnt,A.cnt,E.table.probability)

w=0.5
#homogeneous optimistic attitude with w=0.5
T.st_prob.A<-apply(ABSF,1,T.st_utinity,w=0.5)
#Probability to select the A-arm based on utinity function of T.st: T.st_utinity()
a<-0.8
#ture success rate of A-arm treatment
b<-0.6
#ture success rate of B-arm treatment
T.table.probability<-table.prob(relist(T.st_prob.A,out$sk),p=c(a,b))
#exact probability per 2X2 table based on utinity function of T.st: T.st_utinity()
grid.target.ls<-dist.2D(Success.cnt,A.cnt,T.table.probability) 
 
zlims<-range(c(grid.exp.ls[[N+1]]),c(grid.target.ls[[N+1]]))
par(bg="white",cex=1,col="black",mar=c(2,3.5,3,1))
Vis_3d(g=grid.exp.ls[[N+1]],title="Homogeneous E.st",zlims=zlims)
Vis_3d(g=grid.target.ls[[N+1]],title="Homogeneous T.st",zlims=zlims)