R語言||干貨！tidyverse包-數據轉換dplyr包進階

R-tidyverse包-數據轉換dplyr包進階

R語言||干貨！tidyverse包-數據轉換dplyr包進階

目錄

[TOC]

簡介

Dplyr（https://dplyr.tidyverse.org/）是一種數據操作語法，提供了一組一致的動詞，幫助我們解決最常見的數據操作，比如行操作（filter、slice、arrange）、列操作（slelect、rename、mutate、relocate）、折疊操作（summarise）、合并table（left_join、right_join、inner_join）。查看包中的所有函數：

library(dplyr)
ls('package:dplyr')
  • mutate：添加新列或者修改已存在列
  • select：選擇列變量
  • filter：依據值過濾行
  • summarise：將多個值降為單個summary
  • arrange：行排序
  • left_join：合并兩個數據框
image.png

mutate函數介紹

常規用法

library(tidyverse) # or library(dplyr)

# 創建新列，一列或多列
starwars %>%
  select(name, mass) %>%
  mutate(
    mass2 = mass * 2,
    mass2_squared = mass2 * mass2, .before=name, # .before和.after設置新列插入的位置
    test = 'test'
  )
  
# 創建新列，按行進行統計
starwars %>% select(name, height, mass) %>%rowwise() %>% mutate(m = mean(c(height, mass)))

# 創建新列，根據數值排名創建列
starwars %>%
  select(name, mass, homeworld) %>%
  group_by(homeworld) %>%
  mutate(rank = min_rank(desc(mass)))
  
# 刪除列，設置NULL刪除
starwars %>%
  select(name, height, mass, homeworld) %>%
  mutate(
    mass = NULL
  )

# 修改列，通過運算符修改
starwars %>%
  select(name, height, mass, homeworld) %>%
  mutate(
    height = height * 0.0328084 # convert to feet
  )

# 修改多列，利用across函數將多列字符串數據轉換為因子
starwars %>%
  select(name, homeworld, species) %>%
  mutate(across(!name, as.factor))

# 保留或刪除列，默認保留所有列
df <- tibble(x = 1, y = 2, a = "a", b = "b")
df %>% mutate(z = x + y, .keep = "all") # the default
df %>% mutate(z = x + y, .keep = "used")
df %>% mutate(z = x + y, .keep = "unused")
df %>% mutate(z = x + y, .keep = "none")

特殊用法

# 增加列列名為變量
col_name <- "new_column"
starwars %>% mutate(!!col_name := mass * height)

# 操作列名為變量
col_names <- c("height", "mass")
starwars %>% mutate(across(all_of(col_names), ~ .x * 2))

# 根據條件增加列
starwars %>% mutate(
  gender = case_when(
    sex == "male" ~ "Male",
    sex == "female" ~ "Female",
    TRUE ~ "Unknown"
  )
)

# 根據條件增加列，針對所有以height開頭的列，值乘10
starwars %>% mutate(across(starts_with("height"), ~.x*10), .keep='used')

# 根據條件修改列
starwars %>% mutate(
  mass = if_else(mass > 100, "Yes", "No")
)

# 根據多列條件創建列
starwars %>% mutate(
  special_condition = if_else(mass > 100 & height > 200, "Yes", "No")
, .keep='used')

select函數介紹

常規用法

library(tidyverse)

# For better printing
iris <- as_tibble(iris)

# 選擇列，單列或者多列
starwars %>% select(homeworld, height, mass)

# 按索引選擇列
starwars %>% select(1:3)

# 按范圍選擇列
starwars %>% select(name:mass)

# 反選，使用!或者-反選
starwars %>% select(!(name:mass))

iris %>% select(-c(Sepal.Length, Petal.Length))

# 按條件反選
iris %>% select(!ends_with("Width"))

# 按照多個條件選擇，交集
iris %>% select(starts_with("Petal") & ends_with("Width"))

# 按照多個條件選擇，并集
iris %>% select(starts_with("Petal") | ends_with("Width"))

# 按照多個條件選擇，并集，其中一個條件反選
iris %>% select(starts_with("Petal") & !ends_with("Width"))

特殊用法

# 選擇列為變量
col_names <- c("Sepal.Length", "Sepal.Width")
iris %>% select(all_of(col_names))

col_names <- c("Sepal.Length", "Sepal.Width", "test") # test不在iris數據中
iris %>% select(one_of(col_names))

filter函數介紹

常規用法

# 根據單個條件選擇行
filter(starwars, species == "Human")
filter(starwars, mass > 1000)

# 根據多個條件選擇行
filter(starwars, hair_color == "none" & eye_color == "black")
filter(starwars, hair_color == "none" | eye_color == "black")
filter(starwars, hair_color == "none", eye_color == "black")

# 取反
filter(starwars, !(hair_color == "none" & eye_color == "black"))

# 根據運算結果選擇行
starwars %>% filter(mass > mean(mass, na.rm = TRUE))
starwars %>% group_by(gender) %>% filter(mass > mean(mass, na.rm = TRUE))

# 列名為變量
vars <- c("mass", "height")
cond <- c(80, 150)
starwars %>%
  filter(
    .data[[vars[[1]]]] > cond[[1]],
    .data[[vars[[2]]]] > cond[[2]]
  )

特殊用法

# 列名為變量
col_name <- "Species"
value_to_filter <- "setosa"
filtered_iris <- iris %>% filter(!!sym(col_name) == value_to_filter)

summarise函數介紹

常規用法

# 統計平均值和總行數
mtcars %>%
  summarise(mean = mean(disp), n = n())

# 統計分組后平均值和總行數
mtcars %>%
  group_by(cyl) %>%
  summarise(mean = mean(disp), n = n())

# 列名是變量
var <- "mass"
summarise(starwars, avg = mean(.data[[var]], na.rm = TRUE))

summarise(starwars, avg = mean(!!sym(var), na.rm = TRUE))

特殊用法

# 分組，合并多行為一行
iris %>%
  group_by(Species) %>% summarise(test=paste(Petal.Length, collapse=','))

# 多個函數
min_max <- list(
  min = ~min(.x, na.rm = TRUE), 
  max = ~max(.x, na.rm = TRUE)
)
starwars %>% summarise(across(where(is.numeric), min_max))

starwars %>% summarise(
  tibble(
    across(where(is.numeric), ~min(.x, na.rm = TRUE), .names = "min_{.col}"),
    across(where(is.numeric), ~max(.x, na.rm = TRUE), .names = "max_{.col}")  
  )
)

arrange函數介紹

常規用法

# 依據多行進行排序
arrange(mtcars, cyl, disp)

# 降序排序
arrange(mtcars, desc(disp))

# 先升序，再降序
arrange(mtcars, mpg, desc(hp))

# <p style="color: red;">分組排序，必須設置.by_group參數為TRUE</p>
mtcars %>% group_by(cyl) %>% arrange(desc(wt),.by_group = TRUE) %>% print(n=32)

# 列名是變量
var <- 'mpg'
arrange(mtcars, !!sym(var))
arrange(mtcars, .data[[var]])

# 選擇多列進行排序
iris %>% arrange(pick(starts_with("Sepal")))
iris %>% arrange(across(starts_with("Sepal"), desc))

left_join函數介紹

install.packages('nycflights13')
library(nycflights13)
flights2 <- flights %>% select(year:day, hour, origin, dest, tailnum, carrier)

# 向左合并，指定按某列合并
flights2 %>% left_join(planes, by = "tailnum")

# 按照交集合并
df1 <- tibble(x = c(1, 2), y = 2:1)
df2 <- tibble(x = c(3, 1), a = 10, b = "a")
df1 %>% inner_join(df2)

# 向右合并
df1 %>% right_join(df2)

# 按照并集合并
df1 %>% full_join(df2)

其它常用函數介紹

rename重命名列

# 修改列名
iris <- as_tibble(iris)
rename(iris, petal_length = Petal.Length)

# 修改多列列名
lookup <- c(pl = "Petal.Length", sl = "Sepal.Length")
rename(iris, all_of(lookup))

# 修改多列列名，如果列名不存在，使用any_of
lookup <- c(lookup, new = "unknown")
try(rename(iris, all_of(lookup)))
rename(iris, any_of(lookup))

# 切換列名大小寫，替換字符
rename_with(iris, toupper)
rename_with(iris, toupper, starts_with("Petal"))
rename_with(iris, ~ tolower(gsub(".", "_", .x, fixed = TRUE)))

# 使用paste修改列名時，設置recycle0 = TRUE防止空選
rename_with(
  iris,
  ~ paste0("prefix_", .x, recycle0 = TRUE),
  starts_with("nonexistent")
)

# rlang包中的set_names
set_names(head(mtcars), paste0(colnames(mtcars), "_foo"))
set_names(head(mtcars), paste0, "_foo")

relocate調整列順序

df <- tibble(a = 1, b = 1, c = 1, d = "a", e = "a", f = "a")

# 默認最前
df %>% relocate(f)

# 指定順序
df %>% relocate(a, .after = c)
df %>% relocate(f, .before = b)
df %>% relocate(a, .after = last_col())

# 指定順序并改名
df %>% relocate(ff = f)

# 條件選擇列
df %>% relocate(where(is.character))
df %>% relocate(any_of(c("a", "e", "i", "o", "u")))

slice選擇行

# 選擇第一行
mtcars %>% slice(1)
# 選擇尾部一行或多行
mtcars %>% slice(n())
mtcars %>% slice(5:n())
# 去除指定行
slice(mtcars, -(1:4))

# 選擇頭部或尾部指定行
mtcars %>% slice_head(n = 5)
mtcars %>% slice_tail(n = 5)

# 排序后選擇指定行
mtcars %>% slice_min(mpg, n = 5)
mtcars %>% slice_max(mpg, n = 5)

# slice_min()和slice_max()可能返回多行，因為排序后值相同，設置with_ties = FALSE，輸出指定行
mtcars %>% slice_min(cyl, n = 1)
mtcars %>% slice_min(cyl, n = 1, with_ties = FALSE)

# 隨機選擇行
mtcars %>% slice_sample(n = 5)
mtcars %>% slice_sample(n = 5, replace = TRUE)

# 隨機選擇行，并設置權重
mtcars %>% slice_sample(weight_by = wt, n = 5)

across選擇列

# 選擇列，并用函數處理
iris %>% mutate(across(c(Sepal.Length, Sepal.Width), round))

iris %>%
       group_by(Species) %>%
       summarise(across(starts_with("Sepal"), ~ mean(.x, na.rm = TRUE)))

# 選擇列，并用函數處理，并定義列名。.col表示原來的列名，.fn表示函數名
iris %>%
       group_by(Species) %>%
       summarise(across(starts_with("Sepal"), list(mean, sd), .names = "{.col}.fn{.fn}"))

# 選擇列，只要某一行滿足條件即保留行
iris %>%
       filter(if_any(ends_with("Width"), ~ . > 4))

case_when條件語句

# 單列
x <- 1:70
     case_when(
       x %% 35 == 0 ~ "fizz buzz",
       x %% 5 == 0 ~ "fizz",
       x %% 7 == 0 ~ "buzz",
       .default = as.character(x)
     )

# 多列
starwars %>%
       mutate(type = case_when(
         height > 200 | mass > 200 ~ "large",
         species == "Droid" ~ "robot",
         .default = "other"
       )) %>%
       pull(type)

pull選擇列

# 選擇列，適用于一列，生成vector，和select函數最大的不同
mtcars %>% pull(cyl)

# 選擇列，生成named vector
starwars %>% pull(height, name)
?著作權歸作者所有,轉載或內容合作請聯系作者
平臺聲明：文章內容（如有圖片或視頻亦包括在內）由作者上傳并發布，文章內容僅代表作者本人觀點，簡書系信息發布平臺，僅提供信息存儲服務。

推薦閱讀更多精彩內容