Extracts dumps of pageviews for Turkish language version of Wikipedia for the month of April 2017 and creates basic graphs

Turkish language wikipedia pageviews.R

library("rvest")
library("tidyverse")
library("lubridate")
library("scales")
Sys.setlocale(category = "LC_TIME", locale = "en_IE")

dumpList <- read_html("https://dumps.wikimedia.org/other/pageviews/2017/2017-04/")

links <- data_frame(filename = html_attr(html_nodes(dumpList, "a"), "href")) %>% # extracting links
    filter(grepl(x = filename, "projectviews")) %>% # keeping only aggregated data by project
    mutate(link = paste0("https://dumps.wikimedia.org/other/pageviews/2017/2017-04/", filename))

# create folder to store files
dir.create("wikipediaStats", showWarnings = FALSE)

# download stats by project

for (i in seq_along(links$link)) {
    if (file.exists(file.path("wikipediaStats", paste0(links$filename[i], ".txt")))==FALSE) {
        download.file(url = links$link[i], destfile = file.path("wikipediaStats", paste0(links$filename[i], ".txt")))
    }
}
    
pageviews <- list.files(path = file.path("wikipediaStats"), full.names = TRUE) %>% 
    map_df(.f = read_delim, delim = " ", col_names = FALSE, trim_ws = TRUE, .id = "id") %>% transmute(id = as.integer(id), source = X1, visits = X3) %>% filter(grepl(x = source, pattern = "^tr"))

pageviews <- pageviews %>% group_by(id) %>% tally(visits) %>% bind_cols(links) %>% mutate(date = gsub(x = filename, replacement = "", pattern = "projectviews-")) %>% select(-filename, -link) %>% mutate(date = lubridate::ymd_hms(date, tz = "GMT")) %>%
    mutate(date = date+(3600*4)) # adjusting by four hours to Istanbul time

ggplot(data = pageviews, mapping = aes(x= date, y = n)) +
    geom_line() + 
    scale_y_continuous(name = "", labels = comma) +
    scale_x_datetime(name = "") +
    theme_minimal() +
    labs(title = "Number of pageviews per hour on Turkish-language Wikipedia projects")

ggsave(filename = "1.png")

ggplot(data = pageviews %>% filter(date>as.POSIXct("2017-04-23")), mapping = aes(x= date, y = n)) +
    geom_line() + 
    scale_y_continuous(name = "", labels = comma) +
    scale_x_datetime(name = "") +
    theme_minimal() +
    labs(title = "Number of pageviews per hour on Turkish-language Wikipedia projects")

ggsave(filename = "2.png")

ggplot(data = pageviews %>% filter(date>as.POSIXct("2017-04-28")), mapping = aes(x= date, y = n)) +
    geom_line() + 
    scale_y_continuous(name = "", labels = comma) +
    scale_x_datetime(name = "") +
    theme_minimal() +
    labs(title = "Number of pageviews per hour on Turkish-language Wikipedia projects")

ggsave(filename = "3.png")


pageviews %>% 
    filter(date>lubridate::ymd("2017-04-23")&date<lubridate::ymd("2017-05-01")) %>% 
    mutate(Hour = hour(date), Day = factor(day(date))) %>% 
    ggplot(mapping = aes(x = Hour, y = n, color = Day)) +
    scale_y_continuous(name = "", labels = comma) +
    scale_x_continuous(name = "Hour of the day") +
    geom_line(size = 1.5) +
    scale_color_brewer(type = "qual") +
    theme_minimal() +
    labs(title = "Number of pageviews per hour on Turkish-language Wikipedia projects", subtitle = "23 April-31 April 2017")

ggsave(filename = "hour.png", width = 8, height = 6)

######
# wikipedia projects as listed in the dump files:

# wikibooks: ".b"
# wiktionary: ".d"
# wikimedia: ".m"
# wikipedia mobile: ".mw"
# wikinews: ".n"
# wikiquote: ".q"
# wikisource: ".s"
# wikiversity: ".v"
# mediawiki: ".w"