作業目的: Data Manipulation and Joining

這份作業希望能夠讓你熟悉於清理資料(cleaning data)，並且利用視覺化的文法(grammar of graphics) 呈現結果。過程中會用到前幾週學過 dplyr 的動詞，以及 lubridate 和 ggplot2。每個小題都請寫下你的原始碼並且畫出圖表。

這次的作業使用 readr 提供的原始資料，主題是台灣的電影票房，有興趣的話可以點進 【讀 + 數據】咦！颱風都不來！水庫預警之歷年水庫水位最低是⋯⋯？ 看一下這篇資料新聞

作業: Data Manipulation and Joining

### 這邊不要動
library(tidyverse)
library(lubridate)

df_reservoir_2020 <- 
  read_csv("data/Lab06/reservoir_data_2020.csv", col_names = F)

df_reservoir_2019 <- 
  read_csv("data/Lab06/reservoir_data_2019.csv", col_names = F)

df_reservoir_2018 <- 
  read_csv("data/Lab06/reservoir_data_2018.csv", col_names = F)

df_reservoir_raw <- df_reservoir_2020 %>%
  bind_rows(df_reservoir_2019) %>%
  bind_rows(df_reservoir_2018 ) %>% 
  rename(name = 1, year = 2, area = 3, town = 4, date = 5, height = 6, value = 7, percentage = 8)

### 給你看資料長這樣
df_reservoir_raw %>% head(5)
#> # A tibble: 5 x 8
#>   name      year area     town                    date  height  value percentage
#>   <chr>    <dbl> <chr>    <chr>                   <chr>  <dbl>  <dbl> <chr>     
#> 1 石門水庫  2020 臺灣北區 桃園市龍潭區、大溪區、復興區… 1/1    242.  1.75e4 88.89 %   
#> 2 西勢水庫  2020 臺灣北區 基隆巿暖暖區            1/1     72.1 4.37e1 100 %     
#> 3 新山水庫  2020 臺灣北區 基隆巿安樂區            1/1     83.8 8.68e2 86.61 %   
#> 4 翡翠水庫  2020 臺灣北區 新北市新店區            1/1    167.  3.09e4 92.24 %   
#> 5 寶山水庫  2020 臺灣北區 新竹縣寶山鄉            1/1    140.  4.32e2 85.79 %

1. dplyr & ggplot2::geom_col()

水庫有多少

A. 各個區域(area)有多少縣市有水庫？
B. 各個區域(area)有多少水庫？
C. 請先分別畫，再想辦法畫在一起！

# 第一步：找縣市
df_reservoir_raw %>% mutate(county = str_sub(town, 1, 3))

# 第二步：以區域跟縣市為單位計算有水庫的縣市、水庫數量
df_reservoir_raw %>% mutate(county = str_sub(town, 1, 3)) %>%
  group_by(area) %>% 
  summarise(n_town = n_distinct(county),
            n_name = n_distinct(name))

# 第三步：畫畫
df_reservoir_agg_area <- df_reservoir_raw %>% mutate(county = str_sub(town, 1, 3)) %>%
  group_by(area) %>% 
  summarise(n_town = n_distinct(county),
            n_name = n_distinct(name))

# A. 畫各個區域(area)有多少縣市有水庫
df_reservoir_agg_area %>%
  # ggplot...
  # theme(text = element_text(family = "Noto Sans CJK TC Medium")) 
  # theme(text = element_text(family = "微軟正黑體"))

# B. 畫各個區域(area)有多少水庫
df_reservoir_agg_area %>%
  # ggplot...
  # theme(text = element_text(family = "Noto Sans CJK TC Medium")) 
  # theme(text = element_text(family = "微軟正黑體"))

# C. 畫在一起
df_reservoir_agg_area %>%
  # ggplot...
  facet_wrap(type ~ .) +
  # theme(text = element_text(family = "Noto Sans CJK TC Medium")) 
  # theme(text = element_text(family = "微軟正黑體"))
  
#> Error: <text>:35:0: unexpected end of input
#> 33:   # theme(text = element_text(family = "微軟正黑體"))
#> 34:   
#>    ^

2. dplyr & ggplot2::geom_col()

台灣本島 2018-2020 02/01 蓄水量各區域蓄水量百分比

A. 台灣本島蓄水量各區域蓄水量百分比的堆疊長條圖
B. 台灣本島蓄水量各區域蓄水量百分比的併排長條圖

# 為什麼看本島而已幹嘛欺負澎湖人
df_reservoir_raw %>%
  filter(date == "2/1") %>% 
  group_by(year, area) %>%
  summarise(value = sum(value))

# 第一步: 計算每年、每區的蓄水量
df_reservoir_raw %>%
  filter(area != "澎湖地區") %>%
  filter(date == "2/1") %>% 
  group_by(year, area) %>%
  summarise(value = sum(value)) %>% ungroup()

# 第二步: 計算每年、每區的蓄水量佔比
df_reservoir_raw %>%
  filter(area != "澎湖地區") %>%
  filter(date == "2/1") %>% 
  group_by(year, area) %>%
  summarise(value = sum(value)) %>% ungroup() %>%
  group_by(year) %>% mutate(per = value/sum(value))

# 第三步：畫畫
df_reservoir_agg_year <- df_reservoir_raw %>%
  filter(area != "澎湖地區") %>%
  filter(date == "2/1") %>% 
  group_by(year, area) %>%
  summarise(value = sum(value)) %>% ungroup() %>%
  group_by(year) %>% mutate(per = value/sum(value))

# A. 台灣本島蓄水量各區域蓄水量百分比的堆疊長條圖
df_reservoir_agg_year %>%
  # ggplot...
  # theme(text = element_text(family = "Noto Sans CJK TC Medium")) 
  # theme(text = element_text(family = "微軟正黑體"))

# B. 台灣本島蓄水量各區域蓄水量百分比的併排長條圖
df_reservoir_agg_year %>%
  # ggplot...
  # theme(text = element_text(family = "Noto Sans CJK TC Medium")) 
  # theme(text = element_text(family = "微軟正黑體"))
#> Error: <text>:42:0: unexpected end of input
#> 40:   # theme(text = element_text(family = "微軟正黑體"))
#> 41: 
#>    ^

3. lubridate & ggplot2::geom_line()

石門水庫蓄水量百分比變化

A. 2019 年每天的蓄水量百分比變化
B. 2018-2020年 1 月的蓄水量百分比變化比較

# 先來看看資料長相
df_reservoir_raw %>% filter(name == "石門水庫") %>% count(year)

# A. 2019 年每天的蓄水量百分比變化
df_reservoir_raw %>% filter(name == "石門水庫") %>% filter(year == 2019) %>%
  ggplot(aes(x = date, y = percentage))

df_reservoir_raw %>% select(date, percentage) %>% head(5)

df_reservoir_raw %>% filter(name == "石門水庫") %>% filter(year == 2019) %>%
  mutate(percentage = str_remove_all(percentage, " |\\%")) %>%
  mutate(percentage = as.numeric(percentage)) %>%
  mutate(percentage = percentage/100) %>%
  mutate(date2 = str_c(year, "/", date)) %>%
  mutate(date2 = as_date(date2)) %>%
  filter(!is.na(date2)) %>%
  # ggplot...
  # theme(text = element_text(family = "Noto Sans CJK TC Medium")) 
  # theme(text = element_text(family = "微軟正黑體"))

# B. 石門水庫蓄水量百分比變化: 想看這三年間 1 月的比較
df_reservoir_raw %>% filter(name == "石門水庫") %>% 
  mutate(percentage = str_remove_all(percentage, " |\\%")) %>%
  mutate(percentage = as.numeric(percentage)) %>%
  mutate(percentage = percentage/100) %>%
  mutate(date2 = str_c(year, "/", date)) %>%
  mutate(date2 = as_date(date2)) %>%
  filter(!is.na(date2)) %>%
  mutate(month = month(date2)) %>%
  filter(month == 1) %>%
  # ggplot...
  # theme(text = element_text(family = "Noto Sans CJK TC Medium")) 
  # theme(text = element_text(family = "微軟正黑體"))

df_reservoir_raw %>% filter(name == "石門水庫") %>% 
  mutate(percentage = str_remove_all(percentage, " |\\%")) %>%
  mutate(percentage = as.numeric(percentage)) %>%
  mutate(percentage = percentage/100) %>%
  mutate(date2 = str_c(year, "/", date)) %>%
  mutate(date2 = as_date(date2)) %>%
  filter(!is.na(date2)) %>%
  mutate(month = month(date2)) %>%
  mutate(day = day(date2)) %>%
  filter(month == 1) %>%
  # ggplot...
  # theme(text = element_text(family = "Noto Sans CJK TC Medium")) 
  # theme(text = element_text(family = "微軟正黑體"))

df_reservoir_raw %>% filter(name == "石門水庫") %>% 
  mutate(percentage = str_remove_all(percentage, " |\\%")) %>%
  mutate(percentage = as.numeric(percentage)) %>%
  mutate(percentage = percentage/100) %>%
  mutate(date2 = str_c(year, "/", date)) %>%
  mutate(date2 = as_date(date2)) %>%
  filter(!is.na(date2)) %>%
  mutate(month = month(date2)) %>%
  mutate(day = day(date2)) %>%
  mutate(year = as.factor(year)) %>%
  filter(month == 1) %>%
  # ggplot...
  # theme(text = element_text(family = "Noto Sans CJK TC Medium")) 
  # theme(text = element_text(family = "微軟正黑體"))
#> Error: <text>:64:0: unexpected end of input
#> 62:   # theme(text = element_text(family = "微軟正黑體"))
#> 63: 
#>    ^