-
Notifications
You must be signed in to change notification settings - Fork 0
/
02_mineFlickrData.Rmd
250 lines (204 loc) · 9.79 KB
/
02_mineFlickrData.Rmd
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
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
---
title: "02_mineFlickrData"
author: "Duc-Quang Nguyen"
date: "14 July 2016"
output: html_document
---
* [hex map](http://unconj.ca/blog/not-all-population-maps-are-boring.html)
```{r setup, include=FALSE}
library(knitr)
library(magrittr)
library(dplyr)
library(leaflet)
library(htmltools)
library(swiMap)
library(swiTheme)
library(swiRcharts)
require(ggplot2)
require(rgdal)
require(rgeos)
require(maptools)
require(classInt)
require(viridis)
generateMapAllPics <- F
extraVisualCheck <- F
generateBunchVarMapsInline <- T
generateBunchDistrib <- F
topN <- 10 # get the topN municipalities
startDate <- as.Date("2014-01-01")
endDate <- as.Date("2016-07-14") #Sys.Date()
freqTime <- "week"
data.file <- paste0("input/", startDate, "_", endDate, "_by", freqTime, "_faved_CH.csv")
picifs <- read.csv(file = data.file, stringsAsFactors = F)
```
```{r visual check by mapping data, echo = F, include=F}
path.ch <- getPathShp('CH', year = 2014)
co <- readOGR(path.ch, layer = 'country')
lakes <- readOGR(path.ch, layer = 'lakes')
mu <- readOGR(path.ch, layer = 'municipalities')
co <- spTransform(co, CRS("+init=epsg:4326"))
co.df <- formatShp(co) %>% select(long, lat, order, hole, id, group, NAME, EINWOHNERZ)
lakes <- spTransform(lakes, CRS("+init=epsg:4326"))
mu <- spTransform(mu, CRS("+init=epsg:4326"))
mu.df <- formatShp(mu) %>%
select(long, lat, order, hole, id, group, BFS_NUMMER, NAME, EINWOHNERZ) %>%
rename(lng = long)
# get the municipalities centroids to label map
idList <- mu@data$BFS_NUMMER
centroids.df <- as.data.frame(coordinates(mu))
names(centroids.df) <- c("lon", "lat")
mu.lab <- cbind(centroids.df, id = idList)
mu.lab$label <- as.character(mu.df[match(idList, mu.df$BFS_NUMMER), 'NAME'])
if(generateMapAllPics) {
basem_url <- 'http://{s}.basemaps.cartocdn.com/dark_all/{z}/{x}/{y}.png'
basem_attribution <- '© <a href="http://www.openstreetmap.org/copyright">OpenStreetMap</a> © <a href="http://cartodb.com/attributions">CartoDB</a>'
popup <- paste0("<strong>", picifs$locality, "</strong><br>", htmlLink(picifs$url, "picture"))
map <- leaflet(height = 900) %>%
addTiles(urlTemplate = basem_url, attribution = basem_attribution) %>%
addCircles(
data = picifs,lng = ~lon, lat = ~lat, radius = 0.8,
stroke = FALSE, fillOpacity = 0.3, color = '#996666', popup = popup
)
save_html(
tags$html(
tags$head(includeHTML("styles.html")),
tags$body(
div(class="graphic", map),
HTML(iframeresizer)
)
), file = "flickr_Swiss_checkAll.html", libdir = "js")
original <- list.files("js", "leaflet.css", full.names = T, recursive = T)
file.copy(list.files(system.file("extdata", package="swiRcharts"), 'leaflet.css', full.names = T), original, overwrite = T)
}
```
```{r map pics to municipalities and compute various indicators, warning = F, message= F, echo = F}
# Find in which polygon each geocoordinates fall in !!!
# https://andybeger.com/2014/03/29/associating-points-with-polygons-in-r/
pig.tmp <- picifs
coordinates(pig.tmp) <- ~ lon + lat
proj4string(pig.tmp) <- proj4string(mu)
data <- cbind(picifs, over(pig.tmp, mu) %>% select(BFS_NUMMER))
if(extraVisualCheck) {
cat("\n There are:")
sum(is.na(data$BFS_NUMMER))
cat("\t pics not mapped to any municipalities. Here is map of them")
ch.map <- ggplot(co.df, aes(x = long, y = lat, group = group)) +
geom_path() + coord_equal() + theme_minimal()
ch.map + geom_point(
data = data %>% filter(is.na(BFS_NUMMER)),
aes(x = lon, y = lat, group = 1), alpha = 0.3, size = 1, colour = "red"
)
}
# discasrd picks outside non matching any Swiss municipalities
data %<>% filter(!is.na(BFS_NUMMER))
write.csv(data, file = "input/all_flickr_pics.csv", row.names = F)
commune.dat <- as.data.frame(loadCommunesCHportraits()) %>%
select(Habitants, `Densité de la population par km²`)
# compute the number of pics per commune
com <- data %>% group_by(BFS_NUMMER) %>%
summarise(`Total number of pictures by municipality` = length(id),
`Total number of picture' views by municipality` = sum(views),
`Average number of picture' views by municipality` = mean(views),
`Total number of favorited pictures by municipality` = sum(faveCount),
`Average number of favorited pictures by municipality` = mean(faveCount)
) %>% ungroup()
com <- cbind(com, commune.dat[match(com$BFS_NUMMER, as.numeric(rownames(commune.dat))),])
com$`total number of favorited / total number of pics by municipality` <- com$`Total number of favorited pictures by municipality` / com$`Total number of pictures by municipality`
com$`total number of views / total number of pics by municipality` <- com$`Total number of picture' views by municipality` / com$`Total number of pictures by municipality`
# per capita
com$`Total pictures by capita & by municipality` <- com$`Total number of pictures by municipality` / com$Habitants
com$`Total picture views by capita & by municipality` <- com$`Total number of picture' views by municipality` / com$Habitants
com$`Average #views by capita and by municipality` <- com$`Average number of picture' views by municipality` / com$Habitants
#
com$`Total favorited pictures by capita & by municipality` <- com$`Total number of favorited pictures by municipality` / com$Habitants
com$`Average favorited pictures by capita & by municipality` <- com$`Average number of favorited pictures by municipality` / com$Habitants
vars <- colnames(com)[which(!colnames(com) %in% c(
'BFS_NUMMER', 'Habitants', 'Densité de la population par km²'))]
write.csv(com, file = "input/flickr_byMunicipalities.csv", row.names = F)
subtitle <- paste0("Based on the analysis of all flickr pictures geolocalised in Switzerland and uploaded between ", startDate, " and ", endDate, " (", nrow(data), " pictures)")
plotbyMuni <- function(mu.df, com = com, var, subtitle = "") {
# cat ("\n", "dealing with: ", var)
dd <- com %>% select(one_of(c('BFS_NUMMER', var)))
colnames(dd)[which(colnames(dd) == var)] <- 'variable'
mun <- suppressMessages(left_join(mu.df, dd))
mun$variable[which(is.na(mun$variable))] <- 0
quantile.interval <- unique(quantile(unique(mun$variable), probs = seq(0, 1, by = 1/10), na.rm = T))
mun$varBin <- cut(mun$variable, breaks = quantile.interval, include.lowest = TRUE, dig.lab = 9)
levels(mun$varBin) <- paste0(gsub(",", " - ", gsub("(\\[|\\]|\\(|\\))", "", levels(mun$varBin))), " ")
topBFS <- com[order(com[,var], decreasing = T),] %>% head(topN * 1) %>%
select(BFS_NUMMER) %>% unlist(use.names = F)
maplab <- filter(mu.lab, id %in% topBFS)
maplab <- maplab[match(topBFS, maplab$id),]
maplab$val <- mun[match(topBFS, mun$BFS_NUMMER),'variable']
maplab$txt <- paste0(1:nrow(maplab), ". ", maplab$label)
mapVar <- ggplot(mun, aes(x = lng, y = lat, group = group)) +
geom_polygon(size = 0 , aes(fill = varBin), color = "#e6e6e6") +
swi_theme(y_gridlines = F, base_size = 14) +
theme(
legend.position = "bottom",
axis.line = element_blank(),
axis.ticks = element_blank(),
axis.title = element_blank(),
axis.text = element_blank()
) + scale_fill_viridis(discrete = T, option="plasma", direction = 1, name = var) +
coord_quickmap() +
labs(title = var, subtitle = subtitle, caption = "source: flickr & swisstopo | swissinfo.ch | @duc_qn") +
geom_polygon(
data = co.df,
aes(x = long, y = lat, group = group),
size = 0.3, alpha = 0, colour = "#666666"
) +
geom_label(data = maplab,
aes(x = lon, y=lat, label = txt, group = 1, size = val),
alpha = 0.6, color = "#262626",
label.size = 0, nudge_y = 0.07,
family = "OpenSans-CondensedLight")
print(mapVar + guides(size = FALSE) + scale_size(range = c(4, 6)))
}
# if(generateVarmap) {
# png(file = "flickrPic_viz_allIndicators_%02d.png", width = 1000, height = 800)
# invisible(sapply(vars, function(v) plotbyMuni(mu.df, com, v, subtitle)))
# dev.off()
# }
```
### The top pics
```{r mine data hist, echo = F}
if(generateBunchDistrib) {
plotDistrib <- function(com, var) {
dd <- com
dd$binN <- cut(dd$n, unique(quantile(dd$n, probs = seq(0, 1, 0.1)), na.rm = T))
dd$binDiffUser <- cut(dd$diffUser, unique(quantile(dd$diffUser, probs = seq(0, 1, 0.1)), na.rm = T))
p1 <- ggplot(data = dd) + geom_histogram(aes_string(var)) + theme_minimal()
print(p1 + ggtitle (var))
print(p1 + facet_wrap(~ binN) + ggtitle (paste0(var, " by ", "bins n pics")))
print(p1 + facet_wrap(~ binDiffUser) + ggtitle (paste0(var, " by ", "bins n diff users")))
}
invisible(sapply(vars, function(v) plotDistrib(com, v)))
}
```
## Show heaps of top flickr municipalities
```{r mine data, results = 'asis', message = F, fig.width = 13, fig.height = 11, echo = F}
getTheTopPics <- function(com, v, n = 5, data) {
cat("\n\n")
cat("#", v, " the top ", n, " pics", "\n")
if(generateBunchVarMapsInline) plotbyMuni(mu.df, com = com, var = v, subtitle)
topN <- com[order(com[,v], decreasing = T),] %>% head(n + 3)
cat("\n")
print(knitr::kable(topN, digits = 2, row.names = F))
cat("\n")
mu.names <- as.character(mu.df[match(topN$BFS_NUMMER, mu.df$BFS_NUMMER), "NAME"])
cat("\n")
bfsn <- topN %>% select(BFS_NUMMER) %>% unlist(use.names = F)
names(bfsn) <- mu.names
idx <- match(data$BFS_NUMMER, bfsn)
invisible(sapply(1:length(bfsn), function(i) {
dd <- data[which(idx == i),c('url', 'faveCount', 'views', 'region', 'title', "lon", "lat")] %>%
arrange(desc(faveCount)) %>% head(n)
cat(paste0("* ", names(bfsn)[i], " - ", unique(dd$region),"\n"))
dd %>% select(-region)
cat(paste0(" + [", dd$title, "](", dd$url, ")", "\tfav: ", dd$faveCount, " \t views: ", dd$views, "(", dd$lon, " ", dd$lat, ')\n'))
}))
cat("\n\n")
}
invisible(sapply(vars, function(v) getTheTopPics(com, v, n = topN, data)))
```