Vis_image.cls: Visulization of Benefits of Homogeneous T.st and E.st...
In ryamada22/SelfDecABP:

Description Usage Arguments Details Value Examples

Visulization of 9 classification after obtained numerical outputs from the above cls-fun(), and coded them with different colors in the triangle image.

1	Vis_image.cls(N.p, index, interval)

`N.p`	patients number N.
`index`	a matrices with 5050 raws (number of a,b pairs) and 3 columns (a,b,and numeric classification integers, 1,2,...,9).
`interval`	the interval of sequence of success rate a/b range from [0,1].

The image is made in the example of this function is based on the simulated data ".RData" on the conditions that w = -0.9995,-0.9990,...,0,0.0005,...,0.9990,0.9995, N= 1,2,.,100, for (a,b) = (a,b)|a,b = 0.01,0.02,.,0.98,0.99,a>=b.

Differences of OSRs were calulated at each condition a,b pair (5050 pairs) and each w values (4001 values). And this dataset ".RData" loaded had been classified via function cls_fun().

"gdata" package used here, and color the blue indicates negatives and the red inicates positives in the trangle conditions of a, b pairs.

We saved the simulated data based on E.st and T.st with data form .RData separately in this package, and we can load two datasets to make the image.

trangle image with contour coded with blue-red color.

load("type.index.out.0.0005.RData", ex <- new.env())
names(ex)
N <- 100
interval<-0.01
ps<-combn(seq(0,1,by=interval),2)
w<-seq(-1,1,by=0.0005)
load("type.index.out.0.0005.RData")
Index.out<-type.index.out

table(Index.out[[N]][,3])
# 1    4    5    6    9 
# 1  828 3831  209  181 
#the table of 9 classification when patient N=100
b<-Index.out[[N]][which(Index.out[[N]][,3]==1),1]
b
a<-Index.out[[N]][which(Index.out[[N]][,3]==1),2]
a
# when true success rate a=1, and b=0, the classification 1 exists. 4 with 828 cases of a.b pairs  indicates the optimist outperformed E.st and 
# 9 with 181 pairs indicates E.st outperfomed T.st

library(gdata)
#visulized the above classifications with red, blue and gray separately.
N.id<-c(10,30,50,100)
par(mfrow=c(1,length(N.id)))
for(i in c(N.id)){
index<-Index.out[[i]]
N.p=i
Vis_image.cls(N.p=N.p,index=index,interval=interval)
}