DANN/runtime_plot.R at main · gogolrq/DANN · GitHub

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
rm(list = ls())
setwd("E:/Dropbox/Revision/Distributed KNN/SC_Revision/code")
options(width = 200)
n.sim = c(60000)
gamma.sim = c(0,0.3)
gamma.sim = 0:8/10
#gamma.sim=c(0,0.1,0.2,0.3,0.4,0.5,0.6,0.7,0.8)
kappa.sim=c(0.8)
name="uniform2"
output = c()
result = c()
for (N in n.sim) {
	for (gamma in gamma.sim) {
	  filename=paste(name,"/data/example_N", N, "_gamma", gamma*10, ".txt", sep = "")
	  if (file.exists(filename)) {
		  tryCatch({
			df = read.table(filename, sep = " ")
			result = rbind(result, df)
		  }, error = function(e) {

		  })
	  }
    }
}

data=result

#data.frame(myhat=myhat,qiaohat=qiaohat,myk1=myk1,qiaok1=qiaok1,mytime=time2,qiaotime=time3,N=N,n1=n1,bayes=bayes,newy=newy, kappa=kappa, gamma=gamma)
#result=data.frame(dhat=dhat, myhat=myhat,qiaohat=qiaohat,nn1hat=nn1hat,dk1=dk1,myk1=myk1,qiaok1=qiaok1,dtime=time1, mytime=time2,qiaotime=time3,nn1time=time4,N=N,n1=n1,bayes=bayes,newy=newy, kappa=kappa, gamma=gamma)

colnames(result)=c("dhat","myhat", "qiaohat","nn1hat","dk1","myk1", "qiaok1","dtime", "mytime", "qiaotime","nn1time", "N", "n1", "bayes", "newy", "kappa", "gamma")
result$myhat=(result$myhat>=1/2)
result$qiaohat=(result$qiaohat>=1/2)
result$nn1hat=(result$nn1hat>=1/2)
result$bayes=(result$bayes>=1/2)


result$myhat=(result$myhat==result$bayes)
result$qiaohat=(result$qiaohat==result$bayes)
result$nn1hat=(result$nn1hat==result$bayes)

colnames(result)=c("DA","DAES", "DK","D1","dk1","myk1", "qiaok1","dtime", "mytime", "qiaotime","nn1time", "N", "n1", "bayes", "newy", "kappa", "gamma")

df=result


output=c()
for (N in n.sim){
  for (kappa in kappa.sim){
    for (gamma in gamma.sim) {
			subdf=df[which(df$N==N  & df$gamma==gamma & df$kappa==kappa),]
			print(mean(subdf$myk1==max(subdf$myk1)))
			re=colMeans(subdf)
			re=c(re,dim(subdf)[1])
			#re=c(n,p,re)
			output=rbind(output,re)
    }

  }
}

print(output, digit=3)

data=data.frame(output)

library("tidyverse")
library("latex2exp")
df=select(data,gamma, dtime, mytime)
colnames(df)=c("gamma","DA","DAES")
df=gather(df,key = "Classifier", value = "time", -gamma)
head(df)
df$time=log(df$time)
df$Classifier=factor(df$Classifier, levels=c("DAES","DA"))


titlename=""
if (grepl("uniform",name)){
  titlename="Equally Split"
}

if (grepl("random",name)){
  titlename="Unequally Split"
}


ps.options(horizontal = F)
ps.options(height=3.5, width=4.5)
postscript(paste("../figure/time_",name,".eps",sep=""))
par(mar=c(2,4,2,2))
p1=ggplot(data = df, mapping = aes(x = gamma, y = time, group = Classifier, color=Classifier, pch=Classifier)) +
  geom_line()+
  geom_point(size = 3)+
  xlab(TeX(r'($\epsilon$)'))+
  ylab("")+
  theme(text=element_text(size=15))+
  theme(axis.title.x=element_text(size=20))+
  theme(plot.title = element_text(size=13))+
  ggtitle(titlename)+
  theme(plot.title = element_text(hjust = 0.5))+
  theme(legend.title=element_blank())
  #scale_x_continuous(labels = paste0(xlab, "k"),
  #                   breaks = 10^3 * xlab)
print(p1)
dev.off()

print(p1)