UNIT3A：資料整理與繪圖套件(dplyr & ggplo2)

---

安裝、載入一些基本的套件

if(!require(dplyr)) install.packages("dplyr")
if(!require(tidyr)) install.packages("tidyr")
if(!require(ggplot2)) install.packages("ggplot2")
if(!require(plotly)) install.packages("plotly")
if(!require(babynames)) install.packages("babynames")

---

【Chapter-4】Case Study: US Babynames

檢視資料

str(babynames)

Classes 'tbl_df', 'tbl' and 'data.frame':   1924665 obs. of  5 variables:
 $ year: num  1880 1880 1880 1880 1880 1880 1880 1880 1880 1880 ...
 $ sex : chr  "F" "F" "F" "F" ...
 $ name: chr  "Mary" "Anna" "Emma" "Elizabeth" ...
 $ n   : int  7065 2604 2003 1939 1746 1578 1472 1414 1320 1288 ...
 $ prop: num  0.0724 0.0267 0.0205 0.0199 0.0179 ...

篩選出男生資料

mbaby = babynames %>% 
  rename(number = n) %>% 
  filter(year %in% seq(1880, 2017, 5), sex=="M")

某些男生名字的『數量』

mbaby %>% 
  filter(name %in% c('Steven', 'Thomas', 'Matthew')) %>%
  arrange(name, year) %>% data.frame %>% 
  ggplot(aes(x=year,y=number,col=name)) +
  geom_line()

某些男生名字的『比例』

mbaby %>%
  group_by(year) %>%
  mutate(year_total = sum(number)) %>%
  ungroup() %>%
  filter(name %in% c('Steven', 'Thomas', 'Matthew')) %>%
  mutate(fraction = number / year_total) %>%
  ggplot(aes(x=year,y=fraction,col=name)) +
  geom_line()

簡化程式

mbaby %>%
  group_by(year) %>%
  mutate(fraction = number / sum(number)) %>%
  filter(name %in% c('Steven', 'Thomas', 'Matthew')) %>%
  ggplot(aes(x=year,y=fraction,col=name)) + 
  geom_line()

熱門男生名字的『比例』

mtop = mbaby %>% group_by(year) %>% top_n(1, number) %>% 
  pull(name) %>% unique()

mbaby %>% group_by(year) %>%
  mutate(fraction = number / sum(number)) %>%
  filter(name %in% mtop) %>%
  ggplot(aes(x=year,y=fraction,col=name)) + 
  geom_line(alpha=0.5) + geom_point(size=0.5) -> g
ggplotly(g)  

熱門女生名字的『比例』

fbaby = babynames %>% 
  rename(number = n) %>% 
  filter(year %in% seq(1880, 2017, 5), sex=="F")

ftop = fbaby %>% group_by(year) %>% top_n(1, number) %>% 
  pull(name) %>% unique()

fbaby %>% group_by(year) %>%
  mutate(fraction = number / sum(number)) %>%
  filter(name %in% ftop) %>%
  ggplot(aes(x=year,y=fraction,col=name)) + 
  geom_line(alpha=0.5) + geom_point(size=0.5) -> g

ggplotly(g)  

---

💡 學習重點：
  ■ 每一份資料都是一個物件：名稱、值、種類、結構
  ■ 每一行程式都是一個運算式：物件、運算符號與功能的組合
  ■ 所謂寫程式：使用運算式定義新物件或改變既有物件的過程
  ■ dplyr套件：提供一套方便『整理資料』的…
    § 物件：tibble
    § 運算符號：%>%
    § 功能：
      。select: 依名稱選擇欄位
      。filter: 依條件選取紀錄
      。mutate: 運算新欄位
      。summarise: 欄位統計
      。group_by: 資料分群
      。…

---