専修大学図書館生田本館の混雑率データを取得するRプログラム

plot_library_congestion.R

library(tidyverse)
library(lubridate)
library(scales)
library(ggthemes)
options(dplyr.width = Inf, scipen = 1, digits = 4)
# 任意のフォントに変更してください
theme_set(theme_fivethirtyeight(base_family = "IPAexGothic", base_size = 12))
theme_update(plot.title = element_text(size = 14), axis.title = element_text())
library(rvest)
library(tibbletime)

# 2023-04-01以前はデータが記録されていないようです
start_date = ymd("2023-04-01", tz = "Asia/Tokyo")
# 任意の終了日を指定してください
end_date <- ymd("2023-10-05", tz = "Asia/Tokyo")
date_range <-  seq.POSIXt(start_date, end_date, by = "1 day")

url_base = "https://opac.lib.senshu-u.ac.jp/iwjs0008opc/graph/honkan/graph_"

df <- data.frame()

for (day in date_range){
    date <- format.Date(day, "%Y%m%d")
    url <- paste0(url_base, date, ".html")
    html <- read_html(url)
    datas <- html %>% html_element(".graph") %>% html_text()
    data <- datas %>% str_match(., "data: \\[.+?\\]") %>% str_replace(., "data: \\[", "") %>% str_replace(., "\\]$", "") %>% strsplit(., ",") %>% unlist() %>% as.numeric()

    times <- seq(ymd_hm(paste(date, "09:30"), tz = "Asia/Tokyo"), ymd_hm(paste(date, "20:30"), tz = "Asia/Tokyo"), by = "hour")

    df_tmp <- data.frame(time = times, data = data)
    df <- rbind.data.frame(df, df_tmp)
}

roll_ave <- rollify(mean, window = 12, na_value = NA)
df <- df %>% mutate(ave = roll_ave(data))

write_excel_csv(df, "library_congestion_data.csv")

ggplot(df, aes(x = time, y = data)) +
    geom_line(aes(colour = "毎時データ"), linewidth = 1.1) +
    geom_line(aes(y = ave, colour = "1日平均データ"), linewidth = 1.1, ) +
    scale_x_datetime(date_breaks = "1 month", date_label = "%m/%d") +
    labs(title = "専修大学図書館生田本館の混雑率の推移", x = "", y = "混雑度 (%)") +
    scale_colour_manual(name = "混雑度", breaks = c("毎時データ", "1日平均データ"), values = c("毎時データ" = "gray80", "1日平均データ" = "#008fd5")) +
    theme(legend.position = "top")