-
Notifications
You must be signed in to change notification settings - Fork 0
/
Copy pathcategory-stats.r
134 lines (110 loc) · 4.73 KB
/
category-stats.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
library(parallel)
library(data.table)
source('analysis/foursquare-analysis.r')
baseFolder <- "results/null-model-3"
N_CORES <- detectCores()
THRESH <- 100
MAX_CI <- 30000
k <- 100
countryFiles <- dir("paises")
categoryStats <- data.frame() # global to save it in the workspace image
locationStats <- data.frame() # global to save it in the workspace image
allCheckIns <- data.frame()
collectStatisticsForRanking <- function() {
for(i in 1:length(countryFiles)) {
#readAndCalc <- function(i) {
f <- sprintf("paises/%s", countryFiles[i])
country <- strsplit(countryFiles[i], ".dat", fixed=T)[[1]]
if(country %in% c("Germany", "France", "Spain", "United-Kingdom",
"United-States", "Brazil", "Mexico",
"United-Arab-Emirates", "Saudi-Arabia", "Kuwait", #"Turkey", # run Turkey manually in R shell
"South-Korea", "Malaysia", "Japan", "Thailand")) {
# c("Brazil", "United States", "Indonesia", "France", "Singapore", "Saudi Arabia", "Russia")
message(country)
ci <- readAndFilterCheckIns(f, THRESH)
ci <- filterSelectedCategories(ci)
ci <- resampleIfTooMuchCheckIns(ci)
message(nrow(ci), " check-ins are going to be analyzed")
if(nrow(ci) > 0) {
allCheckIns <<- rbindlist(list(allCheckIns, ci))
stats <- calculateStats(ci, country)
categoryStats <<- rbindlist( list(categoryStats, stats$categoryStats))
locationStats <<- rbindlist( list(locationStats, stats$locationStats))
}
}
}
save.image()
print(categoryStats)
write.table(locationStats, sprintf("%s/location-stats-15-countries-5-categories.csv", baseFolder),
sep="\t", row.names=FALSE)
write.table(categoryStats, sprintf("%s/category-stats-15-countries-5-categories.csv", baseFolder),
sep="\t", row.names=FALSE)
write.table(allCheckIns, sprintf("%s/cleaned-check-ins-1000-15-countries-5-categories.csv", baseFolder),
sep="\t", row.names=FALSE)
}
subcategorySegregationPlots <- function() {
for(country in c("Germany", "France", "Spain", "United-Kingdom",
"United-States", "Brazil", "Mexico",
"United-Arab-Emirates", "Saudi-Arabia", "Kuwait", "Turkey",
"South-Korea", "Malaysia", "Japan", "Thailand")) {
message(country)
f <- sprintf("paises/%s.dat", country)
ci <- readAndFilterCheckIns(f, THRESH)
ci <- filterSelectedCategories(ci)
ci <- resampleIfTooMuchCheckIns(ci)
message(nrow(ci), " check-ins are going to be analyzed")
segregation(ci)
pdf(sprintf("%s/%s/gender-permutation/segregation-subcategories.pdf", baseFolder, country))
segregationData <- segregationSubcategories(ci)
dev.off()
write.table(locationStats, sprintf("%s/%s/gender-permutation/segregation-subcategories.csv",
baseFolder, country), sep="\t", row.names=FALSE)
}
}
##############
# Calculation
#############
calculateStats <- function(ci, region) {
folderName <- sprintf("%s/%s/gender-permutation", baseFolder, region)
generated <- runPermutate(ci, folderName, "permutate-gender", region, k=k, forceGenerate=T)
# segregation() is crucial, the others need male and female popularity,
ci <- segregation(ci, region, log=F)
locationStats <- testObservationWithNullModel(ci, generated, folderName, region, k, PLOT_ANOM_DIST=T)
categoryStats <- getBootstrappedStatistics(folderName, ci, generated, k, region)
return(list(categoryStats=categoryStats$bootstrapStats, locationStats=locationStats))
}
resampleIfTooMuchCheckIns <- function(ci) {
n <- nrow(ci)
if(n > MAX_CI) {
message("Too many check-ins")
ci <- ci[sample(n, MAX_CI, replace=FALSE)]
message("resampled check-ins")
stopifnot(nrow(ci) <= MAX_CI)
}
return(ci)
}
##############
# Run
#############
collectStatisticsForRanking()
#####################################
# Terminal for Turkey
#####################################
country <- "Turkey"
ci <- readAndFilterCheckIns("paises/Turkey.dat", THRESH)
ci <- filterSelectedCategories(ci)
ci <- resampleIfTooMuchCheckIns(ci)
allCheckIns <- fread("results/null-model-4/cleaned-check-ins-1000-15-countries-5-categories.csv")
allCheckIns <<- rbindlist(list(allCheckIns, ci))
stats <- calculateStats(ci, country)
categoryStats <- fread("results/null-model-4/category-stats-15-countries-5-categories.csv")
categoryStats <<- rbindlist(list(categoryStats, stats$categoryStats))
locationStats <- fread("results/null-model-4/location-stats-15-countries-5-categories.csv")
locationStats <<- rbindlist( list(locationStats, stats$locationStats))
save.image()
write.table(locationStats, sprintf("%s/location-stats-15-countries-5-categories.csv", baseFolder),
sep="\t", row.names=FALSE)
write.table(categoryStats, sprintf("%s/category-stats-15-countries-5-categories.csv", baseFolder),
sep="\t", row.names=FALSE)
write.table(allCheckIns, sprintf("%s/cleaned-check-ins-1000-15-countries-5-categories.csv", baseFolder),
sep="\t", row.names=FALSE)