基本動詞
library(dplyr)
library(nycflights13)
dplyr 中最常用的幾個動詞用於修改資料集。
選擇
從資料幀 planes 中選擇 tailnum，type，model 變數：
select(planes, tailnum, type, model)
## # A tibble: 3,322 × 3
## tailnum type model
## <chr> <chr> <chr>
## 1 N10156 Fixed wing multi engine EMB-145XR
## 2 N102UW Fixed wing multi engine A320-214
## 3 N103US Fixed wing multi engine A320-214
## 4 N104UW Fixed wing multi engine A320-214
## 5 N10575 Fixed wing multi engine EMB-145LR
## 6 N105UW Fixed wing multi engine A320-214
## 7 N107US Fixed wing multi engine A320-214
## 8 N108UW Fixed wing multi engine A320-214
## 9 N109UW Fixed wing multi engine A320-214
## 10 N110UW Fixed wing multi engine A320-214
## # ... with 3,312 more rows
使用 magrittr 包中的 forward-pipe 運算子（%>%）重寫上面的語句：
planes %>% select(tailnum, type, model)
## # A tibble: 3,322 × 3
## tailnum type model
## <chr> <chr> <chr>
## 1 N10156 Fixed wing multi engine EMB-145XR
## 2 N102UW Fixed wing multi engine A320-214
## 3 N103US Fixed wing multi engine A320-214
## 4 N104UW Fixed wing multi engine A320-214
## 5 N10575 Fixed wing multi engine EMB-145LR
## 6 N105UW Fixed wing multi engine A320-214
## 7 N107US Fixed wing multi engine A320-214
## 8 N108UW Fixed wing multi engine A320-214
## 9 N109UW Fixed wing multi engine A320-214
## 10 N110UW Fixed wing multi engine A320-214
## # ... with 3,312 more rows
過濾
filter 行基於 crieria。
返回 manufacturer 為 EMBRAER 的資料集：
planes %>% filter(manufacturer == "EMBRAER")
## # A tibble: 299 × 9
## tailnum year type manufacturer model engines
## <chr> <int> <chr> <chr> <chr> <int>
## 1 N10156 2004 Fixed wing multi engine EMBRAER EMB-145XR 2
## 2 N10575 2002 Fixed wing multi engine EMBRAER EMB-145LR 2
## 3 N11106 2002 Fixed wing multi engine EMBRAER EMB-145XR 2
## 4 N11107 2002 Fixed wing multi engine EMBRAER EMB-145XR 2
## 5 N11109 2002 Fixed wing multi engine EMBRAER EMB-145XR 2
## 6 N11113 2002 Fixed wing multi engine EMBRAER EMB-145XR 2
## 7 N11119 2002 Fixed wing multi engine EMBRAER EMB-145XR 2
## 8 N11121 2003 Fixed wing multi engine EMBRAER EMB-145XR 2
## 9 N11127 2003 Fixed wing multi engine EMBRAER EMB-145XR 2
## 10 N11137 2003 Fixed wing multi engine EMBRAER EMB-145XR 2
## # ... with 289 more rows, and 3 more variables: seats <int>, speed <int>,
## # engine <chr>
返回一個資料集，其中 manufacturer 為 EMBRAER，model 為“EMB-145XR”：
planes %>%
filter(manufacturer == "EMBRAER", model == "EMB-145XR")
## # A tibble: 104 × 9
## tailnum year type manufacturer model engines
## <chr> <int> <chr> <chr> <chr> <int>
## 1 N10156 2004 Fixed wing multi engine EMBRAER EMB-145XR 2
## 2 N11106 2002 Fixed wing multi engine EMBRAER EMB-145XR 2
## 3 N11107 2002 Fixed wing multi engine EMBRAER EMB-145XR 2
## 4 N11109 2002 Fixed wing multi engine EMBRAER EMB-145XR 2
## 5 N11113 2002 Fixed wing multi engine EMBRAER EMB-145XR 2
## 6 N11119 2002 Fixed wing multi engine EMBRAER EMB-145XR 2
## 7 N11121 2003 Fixed wing multi engine EMBRAER EMB-145XR 2
## 8 N11127 2003 Fixed wing multi engine EMBRAER EMB-145XR 2
## 9 N11137 2003 Fixed wing multi engine EMBRAER EMB-145XR 2
## 10 N11140 2003 Fixed wing multi engine EMBRAER EMB-145XR 2
## # ... with 94 more rows, and 3 more variables: seats <int>, speed <int>,
## # engine <chr>
上述陳述與寫 AND 條件相同。
planes %>% filter(manufacturer == "EMBRAER" & model == "EMB-145XR")
## # A tibble: 104 × 9
## tailnum year type manufacturer model engines
## <chr> <int> <chr> <chr> <chr> <int>
## 1 N10156 2004 Fixed wing multi engine EMBRAER EMB-145XR 2
## 2 N11106 2002 Fixed wing multi engine EMBRAER EMB-145XR 2
## 3 N11107 2002 Fixed wing multi engine EMBRAER EMB-145XR 2
## 4 N11109 2002 Fixed wing multi engine EMBRAER EMB-145XR 2
## 5 N11113 2002 Fixed wing multi engine EMBRAER EMB-145XR 2
## 6 N11119 2002 Fixed wing multi engine EMBRAER EMB-145XR 2
## 7 N11121 2003 Fixed wing multi engine EMBRAER EMB-145XR 2
## 8 N11127 2003 Fixed wing multi engine EMBRAER EMB-145XR 2
## 9 N11137 2003 Fixed wing multi engine EMBRAER EMB-145XR 2
## 10 N11140 2003 Fixed wing multi engine EMBRAER EMB-145XR 2
## # ... with 94 more rows, and 3 more variables: seats <int>, speed <int>,
## # engine <chr>
對 OR 條件使用管道（|）字元：
planes %>% filter(manufacturer == "EMBRAER" | model == "EMB-145XR")
## # A tibble: 299 × 9
## tailnum year type manufacturer model engines
## <chr> <int> <chr> <chr> <chr> <int>
## 1 N10156 2004 Fixed wing multi engine EMBRAER EMB-145XR 2
## 2 N10575 2002 Fixed wing multi engine EMBRAER EMB-145LR 2
## 3 N11106 2002 Fixed wing multi engine EMBRAER EMB-145XR 2
## 4 N11107 2002 Fixed wing multi engine EMBRAER EMB-145XR 2
## 5 N11109 2002 Fixed wing multi engine EMBRAER EMB-145XR 2
## 6 N11113 2002 Fixed wing multi engine EMBRAER EMB-145XR 2
## 7 N11119 2002 Fixed wing multi engine EMBRAER EMB-145XR 2
## 8 N11121 2003 Fixed wing multi engine EMBRAER EMB-145XR 2
## 9 N11127 2003 Fixed wing multi engine EMBRAER EMB-145XR 2
## 10 N11137 2003 Fixed wing multi engine EMBRAER EMB-145XR 2
## # ... with 289 more rows, and 3 more variables: seats <int>, speed <int>,
## # engine <chr>
將 grepl 與 filter 結合使用可用於圖案匹配條件。
planes %>% filter(grepl("^172.", model))
## # A tibble: 3 × 9
## tailnum year type manufacturer model engines seats
## <chr> <int> <chr> <chr> <chr> <int> <int>
## 1 N378AA 1963 Fixed wing single engine CESSNA 172E 1 4
## 2 N621AA 1975 Fixed wing single engine CESSNA 172M 1 4
## 3 N737MQ 1977 Fixed wing single engine CESSNA 172N 1 4
## # ... with 2 more variables: speed <int>, engine <chr>
之間
返回 year 是 2004 年和 2005 年的所有行：
planes %>% filter(between(year, 2004, 2005))
## # A tibble: 354 × 9
## tailnum year type manufacturer model engines
## <chr> <int> <chr> <chr> <chr> <int>
## 1 N10156 2004 Fixed wing multi engine EMBRAER EMB-145XR 2
## 2 N11155 2004 Fixed wing multi engine EMBRAER EMB-145XR 2
## 3 N11164 2004 Fixed wing multi engine EMBRAER EMB-145XR 2
## 4 N11165 2004 Fixed wing multi engine EMBRAER EMB-145XR 2
## 5 N11176 2004 Fixed wing multi engine EMBRAER EMB-145XR 2
## 6 N11181 2005 Fixed wing multi engine EMBRAER EMB-145XR 2
## 7 N11184 2005 Fixed wing multi engine EMBRAER EMB-145XR 2
## 8 N11187 2005 Fixed wing multi engine EMBRAER EMB-145XR 2
## 9 N11189 2005 Fixed wing multi engine EMBRAER EMB-145XR 2
## 10 N11191 2005 Fixed wing multi engine EMBRAER EMB-145XR 2
## # ... with 344 more rows, and 3 more variables: seats <int>, speed <int>,
## # engine <chr>
切片
slice 僅返回給定索引的行。
返回前五行資料（與基本 head 函式相同）：
planes %>% slice(1:5)
## # A tibble: 5 × 9
## tailnum year type manufacturer model engines
## <chr> <int> <chr> <chr> <chr> <int>
## 1 N10156 2004 Fixed wing multi engine EMBRAER EMB-145XR 2
## 2 N102UW 1998 Fixed wing multi engine AIRBUS INDUSTRIE A320-214 2
## 3 N103US 1999 Fixed wing multi engine AIRBUS INDUSTRIE A320-214 2
## 4 N104UW 1999 Fixed wing multi engine AIRBUS INDUSTRIE A320-214 2
## 5 N10575 2002 Fixed wing multi engine EMBRAER EMB-145LR 2
## # ... with 3 more variables: seats <int>, speed <int>, engine <chr>
返回第 1 行，第 3 行和第 5 行資料：
planes %>% slice(c(1, 3, 5))
## # A tibble: 3 × 9
## tailnum year type manufacturer model engines
## <chr> <int> <chr> <chr> <chr> <int>
## 1 N10156 2004 Fixed wing multi engine EMBRAER EMB-145XR 2
## 2 N103US 1999 Fixed wing multi engine AIRBUS INDUSTRIE A320-214 2
## 3 N10575 2002 Fixed wing multi engine EMBRAER EMB-145LR 2
## # ... with 3 more variables: seats <int>, speed <int>, engine <chr>
返回第一行和最後一行：
planes %>% slice(c(1, nrow(planes)))
## # A tibble: 2 × 9
## tailnum year type manufacturer
## <chr> <int> <chr> <chr>
## 1 N10156 2004 Fixed wing multi engine EMBRAER
## 2 N999DN 1992 Fixed wing multi engine MCDONNELL DOUGLAS CORPORATION
## # ... with 5 more variables: model <chr>, engines <int>, seats <int>,
## # speed <int>, engine <chr>
變異
mutate 可以新增新變數或修改現有變數。
新增一個虛擬變數 engine.dummy，預設值為 0：
planes %>%
mutate(engine.dummy = 0) %>%
select(engine, engine.dummy)
## # A tibble: 3,322 × 2
## engine engine.dummy
## <chr> <dbl>
## 1 Turbo-fan 0
## 2 Turbo-fan 0
## 3 Turbo-fan 0
## 4 Turbo-fan 0
## 5 Turbo-fan 0
## 6 Turbo-fan 0
## 7 Turbo-fan 0
## 8 Turbo-fan 0
## 9 Turbo-fan 0
## 10 Turbo-fan 0
## # ... with 3,312 more rows
使用 dplyr::if_else，如果 engine ==“Turbo-fan”，則將 engine.dummy 設定為 1，否則將 engine.dummy 設定為 0：
planes %>%
mutate(engine.dummy = if_else(engine == "Turbo-fan", 1, 0)) %>%
select(engine, engine.dummy)
## # A tibble: 3,322 × 2
## engine engine.dummy
## <chr> <dbl>
## 1 Turbo-fan 1
## 2 Turbo-fan 1
## 3 Turbo-fan 1
## 4 Turbo-fan 1
## 5 Turbo-fan 1
## 6 Turbo-fan 1
## 7 Turbo-fan 1
## 8 Turbo-fan 1
## 9 Turbo-fan 1
## 10 Turbo-fan 1
## # ... with 3,312 more rows
將 planes$engine 轉換為一個因子。
planes %>%
mutate(engine = as.factor(engine)) %>%
select(engine)
## # A tibble: 3,322 × 1
## engine
## <fctr>
## 1 Turbo-fan
## 2 Turbo-fan
## 3 Turbo-fan
## 4 Turbo-fan
## 5 Turbo-fan
## 6 Turbo-fan
## 7 Turbo-fan
## 8 Turbo-fan
## 9 Turbo-fan
## 10 Turbo-fan
## # ... with 3,312 more rows
安排
使用 arrange 對資料幀進行排序。
由 year 安排 planes：
planes %>% arrange(year)
## # A tibble: 3,322 × 9
## tailnum year type manufacturer model engines
## <chr> <int> <chr> <chr> <chr> <int>
## 1 N381AA 1956 Fixed wing multi engine DOUGLAS DC-7BF 4
## 2 N201AA 1959 Fixed wing single engine CESSNA 150 1
## 3 N567AA 1959 Fixed wing single engine DEHAVILLAND OTTER DHC-3 1
## 4 N378AA 1963 Fixed wing single engine CESSNA 172E 1
## 5 N575AA 1963 Fixed wing single engine CESSNA 210-`5(205)` 1
## 6 N14629 1965 Fixed wing multi engine BOEING 737-524 2
## 7 N615AA 1967 Fixed wing multi engine BEECH 65-A90 2
## 8 N425AA 1968 Fixed wing single engine PIPER PA-28-180 1
## 9 N383AA 1972 Fixed wing multi engine BEECH E-90 2
## 10 N364AA 1973 Fixed wing multi engine CESSNA 310Q 2
## # ... with 3,312 more rows, and 3 more variables: seats <int>,
## # speed <int>, engine <chr>
arrange planes 來自 year desc：
planes %>% arrange(desc(year))
## # A tibble: 3,322 × 9
## tailnum year type manufacturer model engines
## <chr> <int> <chr> <chr> <chr> <int>
## 1 N150UW 2013 Fixed wing multi engine AIRBUS A321-211 2
## 2 N151UW 2013 Fixed wing multi engine AIRBUS A321-211 2
## 3 N152UW 2013 Fixed wing multi engine AIRBUS A321-211 2
## 4 N153UW 2013 Fixed wing multi engine AIRBUS A321-211 2
## 5 N154UW 2013 Fixed wing multi engine AIRBUS A321-211 2
## 6 N155UW 2013 Fixed wing multi engine AIRBUS A321-211 2
## 7 N156UW 2013 Fixed wing multi engine AIRBUS A321-211 2
## 8 N157UW 2013 Fixed wing multi engine AIRBUS A321-211 2
## 9 N198UW 2013 Fixed wing multi engine AIRBUS A321-211 2
## 10 N199UW 2013 Fixed wing multi engine AIRBUS A321-211 2
## # ... with 3,312 more rows, and 3 more variables: seats <int>,
## # speed <int>, engine <chr>
group_by
group_by 允許你通過子集對資料幀執行操作，而無需提取子集。
df <- planes %>% group_by(manufacturer, model)
返回的資料框可能不會顯示為分組。但是，資料幀的 class 和 attributes 將確認它。
class(df)
## [1] "grouped_df" "tbl_df" "tbl" "data.frame"
attributes(df)$vars
## [[1]]
## manufacturer
##
## [[2]]
## model
head(attributes(df)$labels, n = 5L)
## manufacturer model
## 1 AGUSTA SPA A109E
## 2 AIRBUS A319-112
## 3 AIRBUS A319-114
## 4 AIRBUS A319-115
## 5 AIRBUS A319-131
如果要在不刪除現有分組元素的情況下將分組元素新增到資料框，請使用 add 引數設定為 TRUE（預設設定為 FALSE）：
df <- df %>% group_by(type, year, add = TRUE)
class(df)
## [1] "grouped_df" "tbl_df" "tbl" "data.frame"
attributes(df)$vars
## [[1]]
## manufacturer
##
## [[2]]
## model
##
## [[3]]
## type
##
## [[4]]
## year
head(attributes(df)$labels, n = 5L)
## manufacturer model type year
## 1 AGUSTA SPA A109E Rotorcraft 2001
## 2 AIRBUS A319-112 Fixed wing multi engine 2002
## 3 AIRBUS A319-112 Fixed wing multi engine 2005
## 4 AIRBUS A319-112 Fixed wing multi engine 2006
## 5 AIRBUS A319-112 Fixed wing multi engine 2007
如果你想刪除分組使用 ungroup。
df <- df %>% ungroup()
class(df)
## [1] "tbl_df" "tbl" "data.frame"
attributes(df)$vars
## NULL
attributes(df)$labels
## NULL
總結
summarise 用於對資料集進行整體計算或按組進行計算。
找到 mean 的 seats per manufacturer？
planes %>%
group_by(manufacturer) %>%
summarise(Mean = mean(seats))
## # A tibble: 35 × 2
## manufacturer Mean
## <chr> <dbl>
## 1 AGUSTA SPA 8.0000
## 2 AIRBUS 221.2024
## 3 AIRBUS INDUSTRIE 187.4025
## 4 AMERICAN AIRCRAFT INC 2.0000
## 5 AVIAT AIRCRAFT INC 2.0000
## 6 AVIONS MARCEL DASSAULT 12.0000
## 7 BARKER JACK L 2.0000
## 8 BEECH 9.5000
## 9 BELL 8.0000
## 10 BOEING 175.1877
## # ... with 25 more rows
summarise 不會返回未明確分組或包含在彙總函式中的變數。如果要新增另一個變數，則必須將其作為謂詞傳遞給 group_by 或 summarise。
planes %>%
group_by(year, manufacturer) %>%
summarise(Mean = mean(seats))
## Source: local data frame [164 x 3]
## Groups: year [?]
##
## year manufacturer Mean
## <int> <chr> <dbl>
## 1 1956 DOUGLAS 102
## 2 1959 CESSNA 2
## 3 1959 DEHAVILLAND 16
## 4 1963 CESSNA 5
## 5 1965 BOEING 149
## 6 1967 BEECH 9
## 7 1968 PIPER 4
## 8 1972 BEECH 10
## 9 1973 CESSNA 6
## 10 1974 CANADAIR LTD 2
## # ... with 154 more rows
改名
rename 一個變數：
planes %>%
rename(Mfr = manufacturer) %>%
names()
## [1] "tailnum" "year" "type" "Mfr" "model" "engines" "seats"
## [8] "speed" "engine"