rm(list = ls(all = TRUE))
library(doBy)
library(ggplot2)
setwd("~/Documents/tesisGMU/survival")
file = "WankaLab-table.csv"
data = read.csv(file, header = T, skip = 6)
names(data)
## [1] "X.run.number." "RuralShare_PopulationGrowth"
## [3] "seeScarcityLevel." "thresholdToMigrate"
## [5] "systems" "glacier_effect"
## [7] "populationGrowth" "rainSeasonReserve"
## [9] "drySeasonShare" "LimitUrbanArea"
## [11] "prop_INmigrants" "Water_Loss"
## [13] "X.step." "moment_conflict"
## [15] "moment_migrate"
data = data[, c(1, 4, 6, 13, 14, 15)]
names(data)[1] = c("SimulationNumber")
names(data)[4] = c("SimulationDuration")
names(data)[5] = c("TimeUntilConflict")
names(data)[6] = c("TimeUntilMigration")
data <- data[order(data$SimulationNumber, data$SimulationDuration), ]
dataMig = data
rownames(dataMig) = NULL
dataMig$censoredMig = NA
dataConf = data
rownames(dataConf) = NULL
dataConf$censoredConf = NA
dataMig$TimeUntilMigration[dataMig$TimeUntilMigration == -1] = NA
dataMig = as.data.frame(summaryBy(TimeUntilMigration ~ SimulationNumber + thresholdToMigrate +
glacier_effect, data = dataMig, FUN = function(x) {
min = min(x, na.rm = T)
}))
names(dataMig)[4] = c("TimeUntilMigration")
dataConf$TimeUntilConflict[dataConf$TimeUntilConflict == -1] = NA
dataConf = as.data.frame(summaryBy(TimeUntilConflict ~ SimulationNumber + thresholdToMigrate +
glacier_effect, data = dataConf, FUN = function(x) {
min = min(x, na.rm = T)
}))
names(dataConf)[4] = c("TimeUntilConflict")
dataConf$TimeUntilConflict[dataConf$TimeUntilConflict == Inf] = 120
dataConf$status = NA
dataConf$status[dataConf$TimeUntilConflict == 120] = 0
dataConf$status[dataConf$TimeUntilConflict < 120] = 1
summary(dataConf)
## SimulationNumber thresholdToMigrate glacier_effect TimeUntilConflict
## Min. : 1 Min. :2 Min. :0.10 Min. : 98
## 1st Qu.:226 1st Qu.:2 1st Qu.:0.10 1st Qu.:115
## Median :450 Median :3 Median :0.15 Median :120
## Mean :450 Mean :3 Mean :0.15 Mean :117
## 3rd Qu.:675 3rd Qu.:4 3rd Qu.:0.20 3rd Qu.:120
## Max. :900 Max. :4 Max. :0.20 Max. :120
## status
## Min. :0.000
## 1st Qu.:0.000
## Median :0.000
## Mean :0.476
## 3rd Qu.:1.000
## Max. :1.000
summaryBy(TimeUntilConflict ~ glacier_effect, data = dataConf, FUN = function(x) {
c(mean = mean(x), sd = sd(x), me = median(x))
})
## glacier_effect TimeUntilConflict.mean TimeUntilConflict.sd
## 1 0.10 117.7 3.857
## 2 0.15 117.4 3.938
## 3 0.20 116.0 4.912
## TimeUntilConflict.me
## 1 120
## 2 120
## 3 118
DissimilarityBox <- ggplot(dataConf, aes(factor(glacier_effect), TimeUntilConflict))
DissimilarityBox + geom_boxplot(fill = "white", colour = "#3366FF") + xlab("Reduction in water balance as glacier retreats") +
ylab("Simulation Duration") + ggtitle("Effect of Glacier on Time Until Conflict")
dataMig$status = NA
dataMig$status[dataMig$TimeUntilMigration == 120] = 0
dataMig$status[dataMig$TimeUntilMigration < 120] = 1
summary(dataMig)
## SimulationNumber thresholdToMigrate glacier_effect TimeUntilMigration
## Min. : 1 Min. :2 Min. :0.10 Min. : 92
## 1st Qu.:226 1st Qu.:2 1st Qu.:0.10 1st Qu.: 99
## Median :450 Median :3 Median :0.15 Median :102
## Mean :450 Mean :3 Mean :0.15 Mean :102
## 3rd Qu.:675 3rd Qu.:4 3rd Qu.:0.20 3rd Qu.:105
## Max. :900 Max. :4 Max. :0.20 Max. :117
## status
## Min. :1
## 1st Qu.:1
## Median :1
## Mean :1
## 3rd Qu.:1
## Max. :1
summaryBy(TimeUntilMigration ~ glacier_effect, data = dataMig, FUN = function(x) {
c(mean = mean(x), sd = sd(x), me = median(x))
})
## glacier_effect TimeUntilMigration.mean TimeUntilMigration.sd
## 1 0.10 103.4 4.025
## 2 0.15 102.1 3.930
## 3 0.20 101.1 4.228
## TimeUntilMigration.me
## 1 103
## 2 102
## 3 100
DissimilarityBox <- ggplot(dataConf, aes(factor(thresholdToMigrate), TimeUntilConflict))
DissimilarityBox + geom_boxplot(fill = "white", colour = "#3366FF") + xlab("Max time (years) a person waits to migrate ") +
ylab("Simulation Duration") + ggtitle("Effect of Threshold To Migrate on Time Until Conflict")
survobj <- with(dataMig, Surv(TimeUntilMigration, status))
summary(survobj)
## time status
## Min. : 92 Min. :1
## 1st Qu.: 99 1st Qu.:1
## Median :102 Median :1
## Mean :102 Mean :1
## 3rd Qu.:105 3rd Qu.:1
## Max. :117 Max. :1
fit0 <- survfit(survobj ~ 1, data = dataMig)
# summary(fit0)
plot(fit0, xlab = "Survival Time in 6-month periods", ylab = "% Simulations without Migration",
yscale = 100, main = "Survival Distribution (Overall)")
fit1 <- survfit(survobj ~ as.factor(thresholdToMigrate), data = dataMig)
# summary(dataMig$thresholdToMigrate) plot the survival distributions by sex
plot(fit1, xlab = "Survival Time in 6-month periods", ylab = "% Simulations without Migration",
yscale = 100, col = c("red", "blue", "green"), main = "Survival Distributions by thresholdToMigrate")
legend("bottomleft", title = "thresholdToMigrate", c("2", "3", "4"), fill = c("red",
"blue", "green"))
fit1 <- survfit(survobj ~ as.factor(glacier_effect), data = dataMig)
# summary(dataMig$thresholdToMigrate) plot the survival distributions by sex
plot(fit1, xlab = "Survival Time in 6-month periods", ylab = "% Simulations without Migration",
yscale = 100, col = c("red", "blue", "green"), main = "Survival Distributions by thresholdToMigrate")
legend("bottomleft", title = "Glacier Effect", c("0.1", "0.15", "0.2"), fill = c("red",
"blue", "green"))
survobj <- with(dataConf, Surv(TimeUntilConflict, status))
summary(survobj)
## time status
## Min. : 98 Min. :0.000
## 1st Qu.:115 1st Qu.:0.000
## Median :120 Median :0.000
## Mean :117 Mean :0.476
## 3rd Qu.:120 3rd Qu.:1.000
## Max. :120 Max. :1.000
fit0 <- survfit(survobj ~ 1, data = dataConf)
# summary(fit0)
plot(fit0, xlab = "Survival Time in 6-month periods", ylab = "% Simulations without Conflict",
yscale = 100, main = "Survival Distribution (Overall)")
fit1 <- survfit(survobj ~ as.factor(thresholdToMigrate), data = dataConf)
# summary(dataMig$thresholdToMigrate) plot the survival distributions by sex
plot(fit1, xlab = "Survival Time in 6-month periods", ylab = "% Simulations without Conflict",
yscale = 100, col = c("red", "blue", "green"), main = "Survival Distributions by thresholdToMigrate")
legend("bottomleft", title = "thresholdToMigrate", c("2", "3", "4"), fill = c("red",
"blue", "green"))
fit1 <- survfit(survobj ~ as.factor(glacier_effect), data = dataConf)
# summary(dataMig$thresholdToMigrate) plot the survival distributions by sex
plot(fit1, xlab = "Survival Time in 6-month periods", ylab = "% Simulations without Conflict",
yscale = 100, col = c("red", "blue", "green"), main = "Survival Distributions by thresholdToMigrate")
legend("bottomleft", title = "Glacier Effect", c("0.1", "0.15", "0.2"), fill = c("red",
"blue", "green"))
