PoolQueue/Analysis.R at master · elofgren/PoolQueue · 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
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
################################################################
# Statistical Analysis and Graphic Code for Pool-Queue Project #
################################################################

# Import Packages
library(vioplot)

### Analysis of Ecology Results ###
## C- and N- Extinction Probability
PoolExtinction <- as.data.frame(read.csv("LVpoolExtinction.csv"))
QueueExtinction <- as.data.frame(read.csv("LVqueueExtinction.csv"))

# Prey
pool_trials <- nrow(PoolExtinction)
queue_trials <- nrow(QueueExtinction)

extinctprey_pool <- sum(PoolExtinction$Prey)
extinctprey_queue <- sum(QueueExtinction$Prey)

PreyExtinctTest <- prop.test(x = c(extinctprey_pool, extinctprey_queue), n = c(pool_trials,queue_trials))
PreyExtinctDiff <- (extinctprey_pool/pool_trials) - (extinctprey_queue/queue_trials)
PreyExtinctCI <- PreyExtinctTest$conf.int
PreyExtinctP <- PreyExtinctTest$p.value

# Predators
extinctpred_pool <- sum(PoolExtinction$Predator)
extinctpred_queue <- sum(QueueExtinction$Predator)

PredExtinctTest <- prop.test(x = c(extinctpred_pool, extinctpred_queue), n = c(pool_trials,queue_trials))
PredExtinctDiff <- (extinctpred_pool/pool_trials) - (extinctpred_queue/queue_trials)
PredExtinctCI <- PredExtinctTest$conf.int
PredExtinctP <- PredExtinctTest$p.value

# N-extinction
extinctN_pool <- sum(PoolExtinction$N)
extinctN_queue <- sum(QueueExtinction$N)

NExtinctTest <- prop.test(x = c(extinctN_pool, extinctN_queue), n = c(pool_trials,queue_trials))
NExtinctDiff <- (extinctN_pool/pool_trials) - (extinctN_queue/queue_trials)
NExtinctCI <- NExtinctTest$conf.int
NExtinctP <- NExtinctTest$p.value

## Mean and Median Species Populations
PoolMeans <- as.data.frame(read.csv("LVpoolMean.csv"))
QueueMeans <- as.data.frame(read.csv("LVqueueMean.csv"))
PoolMeans$Method <- factor("Pool")
QueueMeans$Method <- factor("Queue")
Means <- rbind(PoolMeans,QueueMeans)
PreyMeansTest <- kruskal.test(Means$Prey,Means$Method)
PredMeansTest <- kruskal.test(Means$Predator,Means$Method)
NMeansTest <- kruskal.test(Means$N,Means$Method)

PoolMedians <- as.data.frame(read.csv("LVpoolMedian.csv"))
QueueMedians <- as.data.frame(read.csv("LVqueueMedian.csv"))
PoolMedians$Method <- factor("Pool")
QueueMedians$Method <- factor("Queue")
Medians <- rbind(PoolMedians,QueueMedians)
PreyMedianTest <- kruskal.test(Medians$Prey,Medians$Method)
PredMedianTest <- kruskal.test(Medians$Predator,Medians$Method)
NMedianTest <- kruskal.test(Medians$N,Medians$Method)

### Analysis of HAI Results ###
HAIPoolExtinction <- as.data.frame(read.csv("CDIPoolExtinction.csv"))
HAIQueueExtinction <- as.data.frame(read.csv("CDIQueueExtinction.csv"))

# N-extinction
haipool_trials <- nrow(HAIPoolExtinction)
haiqueue_trials <- nrow(HAIQueueExtinction)

haiextinct_pool <- sum(HAIPoolExtinction$Extinction)
haiextinct_queue <- sum(HAIQueueExtinction$Extinction)

HAIExtinctTest <- prop.test(x = c(haiextinct_pool, haiextinct_queue), n = c(haipool_trials,haiqueue_trials))
HAIExtinctDiff <- (haiextinct_pool/haipool_trials) - (haiextinct_queue/haiqueue_trials)
HAIExtinctCI <- HAIExtinctTest$conf.int
HAIExtinctP <- HAIExtinctTest$p.value

# Outcomes
HAIPoolOutcomes <- as.data.frame(read.csv("CDIPoolOutcomes.csv"))
HAIQueueOutcomes <- as.data.frame(read.csv("CDIQueueOutcomes.csv"))
HAIPoolOutcomes$Method <- factor("Pool")
HAIQueueOutcomes$Method <- factor("Queue")
HAIOutcomes <- rbind(HAIPoolOutcomes,HAIQueueOutcomes)
IncidentTest <- kruskal.test(HAIOutcomes$Incident,HAIOutcomes$Method)
RecurTest <- kruskal.test(HAIOutcomes$Recur,HAIOutcomes$Method)

### Analysis of Computation Time ###
CompTime <- as.data.frame(read.csv("RunTimes.csv"))
LVComp <- CompTime[which(CompTime$Model=="LV"),]
CDIComp <- CompTime[which(CompTime$Model=="CDI"),]
LVCompTest <- kruskal.test(LVComp$Time ~ LVComp$Method)
CDICompTest <- kruskal.test(CDIComp$Time ~ CDIComp$Method)

### Plot Code ###

## Figures 2 & 3 - Trajectories for predators and prey
PoolPreyTrajectory <- read.csv(("PoolPrey.csv"),header=F)
PPreyT <- as.matrix(PoolPreyTrajectory)
PoolPredTrajectory <- read.csv(("PoolPredator.csv"),header=F)
PPredT <- as.matrix(PoolPredTrajectory)
QueuePreyTrajectory <- read.csv(("QueuePrey.csv"),header=F)
QPreyT <- as.matrix(QueuePreyTrajectory)
QueuePredTrajectory <- read.csv(("QueuePredator.csv"),header=F)
QPredT <- as.matrix(QueuePredTrajectory)
Deterministic <- read.csv(("LVDeterministic.csv"),header=T)

sim_length <- 99 # Number of observations in timeseries - 1
realizations <- 250 # How many realizations of the model were run

par(mfrow=c(1,1))
# Plot Trajectories for Prey
plot(0:sim_length, rep(NA,sim_length+1),main="",xlab="Time", ylab="Prey", ylim=c(0,11.5))
for (i in 1:realizations) {
  lines(0:sim_length, PPreyT [ ,i],lwd=2,col=rgb(0,100,0,20,maxColorValue=255))
}
for (k in 1:realizations) {
  lines(0:sim_length, QPreyT [ ,k],lwd=2,col=rgb(0,0,255,20,maxColorValue=255))
}
lines(Deterministic$Time,Deterministic$prey,lwd=4,col="Black",lty=2)
legend("top",c("Pool","Queue","Deterministic"),lwd=c(2,2,3),col=c("Dark Green","Blue","Black"),lty=c(1,1,2),bty='n', horiz=T)

# Plot Trajectories for Predators
plot(0:sim_length, rep(NA,sim_length+1),main="",xlab="Time", ylab="Predators", ylim=c(0,17))
for (i in 1:realizations) {
  lines(0:sim_length, PPredT [ ,i],lwd=2,col=rgb(0,100,0,20,maxColorValue=255))
}
for (k in 1:realizations) {
  lines(0:sim_length, QPredT [ ,k],lwd=2,col=rgb(0,0,255,20,maxColorValue=255))
}
lines(Deterministic$Time,Deterministic$predator,lwd=4,col="Black",lty=2)
legend("top",c("Pool","Queue","Deterministic"),lwd=c(2,2,3),col=c("Dark Green","Blue","Black"),lty=c(1,1,2),bty='n', horiz=T)

## Figure 4 - Cumultive incident and recurrent cases
par(mfrow=c(1,2))
vioplot(HAIPoolOutcomes$Recur,HAIQueueOutcomes$Recur,names=c("Pool","Queue"),col="Grey",drawRect=FALSE)
title("Recurrent C. difficile", ylab="Cases")
vioplot(HAIPoolOutcomes$Incident,HAIQueueOutcomes$Incident,names=c("Pool","Queue"), col="Grey",drawRect=FALSE)
title("Incident C. difficile", ylab="Cases")

## Figure 5 - Computational runtimes
LVCompP <- subset(LVComp, Method=="Pool")$Time
LVCompQ <- subset(LVComp, Method=="Queue")$Time
CDICompP <- subset(CDIComp, Method=="Pool")$Time
CDICompQ <- subset(CDIComp, Method=="Queue")$Time

par(mfrow=c(1,2))
vioplot(LVCompP,LVCompQ,names=c("Pool","Queue"),col="Grey",drawRect=FALSE)
title("Predator-Prey Model",ylab="Runtime (Seconds)")
vioplot(CDICompP,CDICompQ, names=c("Pool","Queue"),col="Grey",drawRect=FALSE,ylim=c(0,8))
title("Hospital Infection Model",ylab="Runtime (Seconds)")