-
Notifications
You must be signed in to change notification settings - Fork 0
/
cov_table_script.r
195 lines (165 loc) · 8.26 KB
/
cov_table_script.r
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
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
# Remove any previous objects in the workspace
rm(list=ls(all=TRUE))
graphics.off()
source("functions_NCAvNLME_2016.r")
library(plyr)
library(ggplot2)
library(reshape2)
# Choose the directory you wish to compile results from
master.dir <- "E:/hscpw-df1/Data1/Jim Hughes/2016"
setwd(master.dir)
file.name <- "cov_collated_bioq_table.csv"
dir.r16 <- "01_MDV"
df.r16 <- read.csv(paste(dir.r16, file.name, sep="/"))
dir.r78 <- "04_SS"
df.r78 <- read.csv(paste(dir.r78, file.name, sep="/"))
dir.r910 <- "06_genKA"
df.r910 <- read.csv(paste(dir.r910, file.name, sep="/"))
dir.r1114 <- "07_noBOV"
df.r1114 <- read.csv(paste(dir.r1114, file.name, sep="/"))
df.all <- rbind(df.r16, df.r78, df.r910, df.r1114)
df.sub1 <- df.all[1:29]
df.sub2 <- df.all[30:43]
nsim <- 500
nbioq <- df.all$IPREDBE*nsim/100
nbioq.m1 <- df.all$M1IPREDBE*df.all$M1NSIM/100
nbioq.m3 <- df.all$M3IPREDBE*df.all$M3NSIM/100
nnonb <- nsim - nbioq
nnonb.m1 <- df.all$M1NSIM - nbioq.m1
nnonb.m3 <- df.all$M3NSIM - nbioq.m3
cnbioq <- df.all$IPREDCM*nsim/100
cnbioq.m1 <- df.all$M1IPREDCM*df.all$M1NSIM/100
cnbioq.m3 <- df.all$M3IPREDCM*df.all$M3NSIM/100
cnnonb <- nsim - cnbioq
cnnonb.m1 <- df.all$M1NSIM - cnbioq.m1
cnnonb.m3 <- df.all$M3NSIM - cnbioq.m3
ncaTP <- nbioq - df.all$NCAFT2*nsim/100
ncaTN <- nnonb - df.all$NCAFT1*nsim/100
cncaTP <- cnbioq - df.all$NCACT2*nsim/100
cncaTN <- cnnonb - df.all$NCACT1*nsim/100
m1f1TP <- nbioq.m1 - df.all$M1F1FT2*df.all$M1NSIM/100
m1f1TN <- nnonb.m1 - df.all$M1F1FT1*df.all$M1NSIM/100
cm1TP <- cnbioq.m1 - df.all$M1CT2*df.all$M1NSIM/100
cm1TN <- cnnonb.m1 - df.all$M1CT1*df.all$M1NSIM/100
m1phTP <- nbioq.m1 - df.all$M1PHFT2*df.all$M1NSIM/100
m1phTN <- nnonb.m1 - df.all$M1PHFT1*df.all$M1NSIM/100
m3f1TP <- nbioq.m3 - df.all$M3F1FT2*df.all$M3NSIM/100
m3f1TN <- nnonb.m3 - df.all$M3F1FT1*df.all$M3NSIM/100
cm3TP <- cnbioq.m3 - df.all$M3CT2*df.all$M3NSIM/100
cm3TN <- cnnonb.m3 - df.all$M3CT1*df.all$M3NSIM/100
m3phTP <- nbioq.m3 - df.all$M3PHFT2*df.all$M3NSIM/100
m3phTN <- nnonb.m3 - df.all$M3PHFT1*df.all$M3NSIM/100
df.sub1$NCAFSENS <- ncaTP/nbioq*100
df.sub1$M1F1FSENS <- m1f1TP/nbioq.m1*100
df.sub1$M1PHFSENS <- m1phTP/nbioq.m1*100
df.sub1$M3F1FSENS <- m3f1TP/nbioq.m3*100
df.sub1$M3PHFSENS <- m3phTP/nbioq.m3*100
df.sub1$NCAFSPEC <- ncaTN/nnonb*100
df.sub1$M1F1FSPEC <- m1f1TN/nnonb.m1*100
df.sub1$M1PHFSPEC <- m1phTN/nnonb.m1*100
df.sub1$M3F1FSPEC <- m3f1TN/nnonb.m3*100
df.sub1$M3PHFSPEC <- m3phTN/nnonb.m3*100
df.sub1$NCAFACC <- (ncaTP+ncaTN)/nsim*100
df.sub1$M1F1FACC <- (m1f1TP+m1f1TN)/df.all$M1NSIM*100
df.sub1$M1PHFACC <- (m1phTP+m1phTN)/df.all$M1NSIM*100
df.sub1$M3F1FACC <- (m3f1TP+m3f1TN)/df.all$M3NSIM*100
df.sub1$M3PHFACC <- (m3phTP+m3phTN)/df.all$M3NSIM*100
df.sub1$NCACSENS <- cncaTP/nbioq*100
df.sub1$M1CSENS <- cm1TP/nbioq.m1*100
df.sub1$M3CSENS <- cm3TP/nbioq.m3*100
df.sub1$NCACSPEC <- cncaTN/nnonb*100
df.sub1$M1CSPEC <- cm1TN/nnonb.m1*100
df.sub1$M3CSPEC <- cm3TN/nnonb.m3*100
df.sub1$NCACACC <- (cncaTP+cncaTN)/nsim*100
df.sub1$M1CACC <- (cm1TP+cm1TN)/df.all$M1NSIM*100
df.sub1$M3CACC <- (cm3TP+cm3TN)/df.all$M3NSIM*100
df.final <- data.frame(df.sub1, df.sub2)
df.final.all <- df.final[df.final$TERMSTAT == "All", ]
df.final.suc <- df.final[df.final$TERMSTAT == "Only Success", ]
write.csv(df.final.suc,"cov_results_successonly.csv",row.names=F)
write.csv(df.final.suc,"cov_results_all.csv",row.names=F)
# PLOT SET 1
#Determine median, upper lower bounds of the difference between using
#all runs and only using successfully minimised runs
df.diff <- data.frame(
M1F1.F.SPEC = df.final.suc$M1F1FSPEC - df.final.all$M1F1FSPEC,
M1PH.F.SPEC = df.final.suc$M1PHFSPEC - df.final.all$M1PHFSPEC,
M1F1.F.SENS = df.final.suc$M1F1FSENS - df.final.all$M1F1FSENS,
M1PH.F.SENS = df.final.suc$M1PHFSENS - df.final.all$M1PHFSENS,
M1F1.F.ACC = df.final.suc$M1F1FACC - df.final.all$M1F1FACC,
M1PH.F.ACC = df.final.suc$M1PHFACC - df.final.all$M1PHFACC,
M1.C.SPEC = df.final.suc$M1CSPEC - df.final.all$M1CSPEC,
M1.C.SENS = df.final.suc$M1CSENS - df.final.all$M1CSENS,
M1.C.ACC = df.final.suc$M1CACC - df.final.all$M1CACC)
#colwise(summary)(df.diff)
df.final.all.M1F1FSPEC <- df.final.all$M1F1FSPEC
df.final.all.M1F1FSPEC[df.final.all.M1F1FSPEC == 0] <- 0.0000001
df.ratio <- data.frame(
M1F1.F.SPEC = df.final.suc$M1F1FSPEC/df.final.all.M1F1FSPEC,
M1PH.F.SPEC = df.final.suc$M1PHFSPEC/df.final.all$M1PHFSPEC,
M1F1.F.SENS = df.final.suc$M1F1FSENS/df.final.all$M1F1FSENS,
M1PH.F.SENS = df.final.suc$M1PHFSENS/df.final.all$M1PHFSENS,
M1F1.F.ACC = df.final.suc$M1F1FACC/df.final.all$M1F1FACC,
M1PH.F.ACC = df.final.suc$M1PHFACC/df.final.all$M1PHFACC,
M1.C.SPEC = df.final.suc$M1CSPEC/df.final.all$M1CSPEC,
M1.C.SENS = df.final.suc$M1CSENS/df.final.all$M1CSENS,
M1.C.ACC = df.final.suc$M1CACC/df.final.all$M1CACC)
#colwise(summary)(df.ratio)
df.diff.l <- melt(df.diff)
df.diff.l <- data.frame(colsplit(df.diff.l$variable, pattern = "\\.",
names = c("method", "bioq", "stat")), value = df.diff.l$value)
df.ratio.l <- melt(df.ratio)
df.ratio.l <- data.frame(colsplit(df.ratio.l$variable, pattern = "\\.",
names = c("method", "bioq", "stat")), value = df.ratio.l$value)
df.diff.l$statf <- factor(df.diff.l$stat)
levels(df.diff.l$statf) <- c("Accuracy", "Sensitivity", "Specificity")
df.ratio.l$statf <- factor(df.ratio.l$stat)
levels(df.ratio.l$statf) <- c("Accuracy", "Sensitivity", "Specificity")
theme_set(theme_bw())
titletext1 <- expression(atop("Change in Results (Accuracy, Sensitivity, Specificity)",
atop("Difference between using only successfully minimised runs and using all runs to determine bioequivalence")))
plotobj1 <- NULL
plotobj1 <- ggplot(data = df.diff.l[df.diff.l$bioq == "F", ])
plotobj1 <- plotobj1 + ggtitle(titletext1)
plotobj1 <- plotobj1 + geom_boxplot(aes(factor(method), value))
plotobj1 <- plotobj1 + scale_x_discrete("\nMethod")
plotobj1 <- plotobj1 + scale_y_continuous("Change in Percentage\n")
plotobj1 <- plotobj1 + facet_wrap(~statf)
plotobj1
ggsave("cov_DiffPlot_F1.png", width=20, height=16, units=c("cm"))
plotobj1b <- plotobj1 + scale_y_continuous("Change in Percentage\n", lim = c(-20, 20))
plotobj1b
ggsave("cov_DiffPlot_F1_lim.png", width=20, height=16, units=c("cm"))
CI90lo <- function(x) quantile(x,probs = 0.05, na.rm = T)
CI90hi <- function(x) quantile(x,probs = 0.95, na.rm = T)
df.diff.stat <- ddply(df.diff.l, .(method, bioq, statf), function(x) {
c(CI90lo(x$value), mean(x$value, na.rm = T), CI90hi(x$value))
})
names(df.diff.stat)[4:6] <- c("ci90lo", "mean", "ci90hi")
titletext1 <- expression(atop("Change in Results (Accuracy, Sensitivity, Specificity)",
atop("CI of the difference between use of covariance step passing runs and all runs for bioequivalence")))
plotobj1a <- NULL
plotobj1a <- ggplot(data = df.diff.stat[df.diff.stat$bioq == "F", ], aes(factor(method), mean))
plotobj1a <- plotobj1a + ggtitle(titletext1)
plotobj1a <- plotobj1a + geom_point()
plotobj1a <- plotobj1a + geom_errorbar(aes(ymin = ci90lo, ymax = ci90hi), width = 0.5)
plotobj1a <- plotobj1a + scale_x_discrete("\nMethod")
plotobj1a <- plotobj1a + scale_y_continuous("Change in Percentage\n",
labels = dollar_format(suffix = "%", prefix = ""), lim = c(-8, 8))
plotobj1a <- plotobj1a + facet_wrap(~statf)
plotobj1a
ggsave("cov_CIDiffPlot_F1_lim.png", width=20, height=8, units=c("cm"))
stat2 <- read.csv("dfdiffstat.csv")
df.diff.stat2 <- stat2[stat2$bioq == "F" & (stat2$method == "M1F1" | stat2$method == "M1PH"),]
plotobj1c <- plotobj1a + geom_point(data = df.diff.stat2, colour = "grey50")
plotobj1c <- plotobj1c + geom_errorbar(aes(ymin = ci90lo, ymax = ci90hi), data = df.diff.stat2, width = 0.5, colour = "grey50")
ggsave("compar_CIDiffPlot_F1.png", width=20, height=8, units=c("cm"))
# plotobj2 <- NULL
# plotobj2 <- ggplot(data = df.diff.l[df.diff.l$bioq == "C", ])
# plotobj2 <- plotobj2 + ggtitle(titletext1)
# plotobj2 <- plotobj2 + geom_boxplot(aes(factor(method), value))
# plotobj2 <- plotobj2 + scale_x_discrete("\nMethod")
# plotobj2 <- plotobj2 + scale_y_continuous("Change in Percentage\n")
# plotobj2 <- plotobj2 + facet_wrap(~statf)
# plotobj2
# ggsave("DiffPlot_CM.png", width=20, height=16, units=c("cm"))