-
Notifications
You must be signed in to change notification settings - Fork 0
/
roraas.R
137 lines (99 loc) · 4.16 KB
/
roraas.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
library(boot)
library(lme4)
library(arm)
library(msm)
library(nlme)
library(plyr) #plyr needs to be loaded before dplyr to avoid conflicts
library(dplyr)
set.seed(seed = 42)
P = 48; S = 10; R = 2; meanMale = 1; meanFemale = 1.5;
CVgFemale = 30; CVi = 10; CVa = 5; CVgMale = 30
Subject <- rep(1:P, each = S * R)
nGender = P*S*R/2
# nMale = P-nFemale
Gender <- c(rep("Female", nGender), rep("Male", nGender))
Time <- rep(1:S, each = R, times = P)
Replicate <- rep(1:R, each = 1, times = P * S)
femaleValues = rep(meanFemale*rtnorm(P/2,1,CVgFemale/100, 1-3*CVgFemale/100, 1+3*CVgFemale/100), each = S*R)
maleValues = rep(meanMale*rtnorm(P/2,1,CVgMale/100, 1-3*CVgMale/100, 1+3*CVgMale/100), each = S*R)
mean(femaleValues)
mean(maleValues)
value1 <- c(femaleValues,maleValues)
value2 <- rep(rtnorm(S*P,1,CVi/100, 1-3*CVi/100, 1+3*CVi/100), each = R)
value3 <- rep(rtnorm(P*S*R,1,CVa/100, 1-3*CVa/100, 1+3*CVa/100), each = 1)
Measurement <- value1 * value2 * value3
output <- (data.frame(Subject, Gender, Time, Replicate, Measurement))
output$Subject <- as.factor(output$Subject)
output$Time <- as.factor(output$Time)
output$Replicate <- as.factor(output$Replicate)
output$Measurement = as.numeric(formatC(output$Measurement, format = "f"))
head(output)
boxplot(output$Measurement ~ output$Gender)
mean = aggregate(output[, "Measurement"], list(output$Gender), mean)
sd = aggregate(output[, "Measurement"], list(output$Gender), sd)
mean
sd
round(((sd$x/mean$x)*100),2)
write.table(output, "~/Documents/Studies/BiologicalVariation/Rcodes/example_data.txt", quote = F, row.names = F, sep = "\t")
test.data<-create.nested.data(P = 24, S = 10, R = 2, pop.mean = 100, CVg = 30, CVi = 10, CVa = 5)
head(test.data)
write.table(output, "~/Documents/Studies/BiologicalVariation/Rcodes//example_data.txt", quote = F, row.names = F, sep = "\t")
inputdata = test.data
normalise.nested.data <- function(inputdata){
tmp <- inputdata %>%
group_by(Subject, Time) %>%
mutate(replicate.count = n(), sample.sum = sum(Measurement, na.rm = TRUE))
output <- tmp %>%
group_by(Subject) %>%
mutate(subject_mean = sum(sample.sum, na.rm =
TRUE)/sum(replicate.count, na.rm = TRUE),
normalised_measurment = Measurement / subject_mean)
return(output)
}
normalise.nested.data(test.data)
test.data$Subject <- as.factor(test.data$Subject)
test.data$Sample <- as.factor(test.data$Sample)
test.data$Replicate <- as.factor(test.data$Replicate)
test.data <- test.data[with(test.data, order(Subject, Sample,Replicate)), ]
head(test.data)
test.data <- na.omit(test.data)
testdatafile<-normalise.nested.data(test.data)
head(testdatafile)
cv.formula <- normalised_measurment ~ (1|Subject) + (1|Subject:Sample)
sd.formula <- measurment ~ (1|Subject) + (1|Subject:Sample)
mod1<-lmer(data=testdatafile, REML = TRUE, cv.formula)
display(mod1)
ci.model <- confint(mod1)
create.new.dataset.indices<-function(inputdata, indices){
output <- {}
for(i in 1:length(indices)){
tmp<-subset(inputdata, inputdata[["Subject"]] == indices[i])
tmp$Subject<-i
output<-rbind(output,tmp)
}
return(output)
}
bootstrap.nested.data <- function(inputdata){
#these nested samples are used to find the variance, and re-sampled
# from the indices from the bootstrap samples
anova_bootstrap_nested <- function(formula, data, indices) {
d <- create.new.dataset.indices(inputdata, indices)
mod<-lmer(data = d, formula)
return(sigma.hat(mod)$sigma$'Subject:Sample')
}
anova_bootstrap <- function(formula, data, indices) {
d <- create.new.dataset.indices(inputdata, indices)
mod<-lmer(data = d, formula)
#we resample from the indices for each resampled dataset for the
# nested sampling important to use the correct subject number throught
nested <- boot(indices, statistic=anova_bootstrap_nested, R=N,
formula = formula)
return(c(sigma.hat(mod)$sigma$'Subject:Sample', var(nested$t)))
}
n.cores <- 32; B<- 1999; N<-999;
results <- boot(unique(inputdata$Subject), statistic=anova_bootstrap,
R=B,
formula = cv.formula, parallel = "multicore", ncpus =
n.cores)
return(results)
}