R語(yǔ)言基礎(chǔ)系列：

• R語(yǔ)言基礎(chǔ)1--R數(shù)據(jù)格式：.rds和.rda的區(qū)別
• R語(yǔ)言基礎(chǔ)2--數(shù)據(jù)排序與長(zhǎng)寬型數(shù)據(jù)的轉(zhuǎn)換
• R語(yǔ)言基礎(chǔ)3--tidyverse包總結(jié)

查看dplyr包中有哪些函數(shù)

library(dplyr)
ls('package:dplyr')
#目前有290個(gè)包

1：篩選函數(shù)

• 1.1 filter函數(shù)?? 針對(duì)行進(jìn)行操作，提取一個(gè)或多個(gè)分組變量中的某個(gè)觀測(cè)

library(dplyr)
library(reshape2) #使用的演示數(shù)據(jù)集來(lái)自這個(gè)包
#選擇tips數(shù)據(jù)框中非吸煙和周日的行進(jìn)行篩選
sub <- filter(tips,tips$smoker=='No',tips$day=='Sun')
head(sub)
#   total_bill  tip    sex smoker day   time size
# 1      16.99 1.01 Female     No Sun Dinner    2
# 2      10.34 1.66   Male     No Sun Dinner    3
# 3      21.01 3.50   Male     No Sun Dinner    3
# 4      23.68 3.31   Male     No Sun Dinner    2
# 5      24.59 3.61 Female     No Sun Dinner    4
# 6      25.29 4.71   Male     No Sun Dinner    4

??filter和%in%結(jié)合使用，可以用于選取一個(gè)列中的多個(gè)分類(lèi)變量

sub1 <- filter(tips,day %in% c('Sat','Sun'))

filter()只能篩選出條件為T(mén)RUE的行，它會(huì)排除那些條件為FALSE和NA的行。
注意：filter函數(shù)約等于subsets，但subsets既可以對(duì)行進(jìn)行操作，也可以對(duì)列進(jìn)行操作。

• 1.2. slice函數(shù) 針對(duì)行進(jìn)行操作，可以提取指定行數(shù)

 sub2 <- slice(tips,1:5) #tips是要操作的數(shù)據(jù)框，1：5是提取的行
 sub2
#  total_bill  tip    sex smoker day   time size
# 1      16.99 1.01 Female     No Sun Dinner    2
# 2      10.34 1.66   Male     No Sun Dinner    3
# 3      21.01 3.50   Male     No Sun Dinner    3
# 4      23.68 3.31   Male     No Sun Dinner    2
# 5      24.59 3.61 Female     No Sun Dinner    4

• 1.3. select函數(shù)?? 針對(duì)列進(jìn)行操作

 sub3 <- select(tips,tip,sex,smoker) #提取tips中的tip, sex, smoker這三列
 head(sub3)
#   tip    sex smoker
# 1 1.01 Female     No
# 2 1.66   Male     No
# 3 3.50   Male     No
# 4 3.31   Male     No
# 5 3.61 Female     No
# 6 4.71   Male     No

 sub4 <- select(tips,2:5) #提取tips中的2-5列
 head(sub4)
#   tip    sex smoker day
# 1 1.01 Female     No Sun
# 2 1.66   Male     No Sun
# 3 3.50   Male     No Sun
# 4 3.31   Male     No Sun
# 5 3.61 Female     No Sun
# 6 4.71   Male     No Sun

 sub5 <- select(tips,tip:time) #提取tips中從tip到time所有的列
 head(sub5)
#   tip    sex smoker day   time
# 1 1.01 Female     No Sun Dinner
# 2 1.66   Male     No Sun Dinner
# 3 3.50   Male     No Sun Dinner
# 4 3.31   Male     No Sun Dinner
# 5 3.61 Female     No Sun Dinner
# 6 4.71   Male     No Sun Dinner

2. arrange函數(shù)（排序函數(shù)）??

 new_tips <- arrange(tips,total_bill,tip) #如果total_bill是一樣的，就按tip排序
 head(new_tips)
#    total_bill  tip    sex smoker  day   time size
# 68        3.07 1.00 Female    Yes  Sat Dinner    1
# 93        5.75 1.00 Female    Yes  Fri Dinner    2
# 112       7.25 1.00 Female     No  Sat Dinner    1
# 173       7.25 5.15   Male    Yes  Sun Dinner    2
# 150       7.51 2.00   Male     No Thur  Lunch    2
# 196       7.56 1.44   Male     No Thur  Lunch    2

根據(jù)total_bill和tips對(duì)數(shù)據(jù)框進(jìn)行排序（默認(rèn)升序）

#降序
 new_tips <- arrange(tips,desc(total_bill),tip)
 head(new_tips)
#    total_bill   tip    sex smoker day   time size
# 171      50.81 10.00   Male    Yes Sat Dinner    3
# 213      48.33  9.00   Male     No Sat Dinner    4
# 60       48.27  6.73   Male     No Sat Dinner    4
# 157      48.17  5.00   Male     No Sun Dinner    6
# 183      45.35  3.50   Male    Yes Sun Dinner    3
# 103      44.30  2.50 Female    Yes Sat Dinner    3

缺失值總是排在最后

3. rename函數(shù)（對(duì)列進(jìn)行重新命名）

 new_tips <- rename(tips,bill=total_bill)
 head(new_tips)
#    bill  tip    sex smoker day   time size
# 1 16.99 1.01 Female     No Sun Dinner    2
# 2 10.34 1.66   Male     No Sun Dinner    3
# 3 21.01 3.50   Male     No Sun Dinner    3
# 4 23.68 3.31   Male     No Sun Dinner    2
# 5 24.59 3.61 Female     No Sun Dinner    4
# 6 25.29 4.71   Male     No Sun Dinner    4

4. distinct函數(shù)（與levels函數(shù)有異曲同工之妙）

levels(tips$sex)
# [1] "Female" "Male"
distinct(tips,sex)
#     sex
# 1 Female
# 2   Male
distinct(tips,day)
#    day
# 1   Sun
# 20  Sat
# 78 Thur
# 91  Fri

5. mutate函數(shù) & transform函數(shù)（生成新的變量）??

head(mutate(tips,rate=tip/total_bill))
#   total_bill  tip    sex smoker day   time size       rate
# 1      16.99 1.01 Female     No Sun Dinner    2 0.05944673
# 2      10.34 1.66   Male     No Sun Dinner    3 0.16054159
# 3      21.01 3.50   Male     No Sun Dinner    3 0.16658734
# 4      23.68 3.31   Male     No Sun Dinner    2 0.13978041
# 5      24.59 3.61 Female     No Sun Dinner    4 0.14680765
# 6      25.29 4.71   Male     No Sun Dinner    4 0.18623962

新生成了rate變量

head(mutate(tips,rate=tip/total_bill,new_rat=rate*100))
#   total_bill  tip    sex smoker day   time size       rate   new_rat
# 1      16.99 1.01 Female     No Sun Dinner    2 0.05944673  5.944673
# 2      10.34 1.66   Male     No Sun Dinner    3 0.16054159 16.054159
# 3      21.01 3.50   Male     No Sun Dinner    3 0.16658734 16.658734
# 4      23.68 3.31   Male     No Sun Dinner    2 0.13978041 13.978041
# 5      24.59 3.61 Female     No Sun Dinner    4 0.14680765 14.680765
# 6      25.29 4.71   Male     No Sun Dinner    4 0.18623962 18.623962

rate和new_rate可以同步生成
transform函數(shù)與mutate函數(shù)的不同之處在于：mutate函數(shù) 可以同時(shí)生成有遞進(jìn)關(guān)系的多個(gè)變量，而 transform函數(shù)只能一個(gè)一個(gè)生成。

head(transform(tips,rate=tip/total_bill,new_rat=rate*100))
# Error in eval(substitute(list(...)), `_data`, parent.frame()) : 
#   object 'rate' not found

transform函數(shù)必須先生成rate再生成new_rate，mutate函數(shù)可以同時(shí)生成rate和new_rate。
如果只想保留新變量，可以使用transmute函數(shù)()。

6. sample_n函數(shù) & sample_frac函數(shù)（在數(shù)據(jù)框中隨機(jī)抽取一些行）

sample_n(iris,size=10) #從iris里隨機(jī)抽取了10行
#    Sepal.Length Sepal.Width Petal.Length Petal.Width    Species
# 1           6.4         2.8          5.6         2.1  virginica
# 2           4.4         3.2          1.3         0.2     setosa
# 3           4.3         3.0          1.1         0.1     setosa
# 4           7.0         3.2          4.7         1.4 versicolor
# 5           5.4         3.0          4.5         1.5 versicolor
# 6           5.4         3.4          1.7         0.2     setosa
# 7           7.6         3.0          6.6         2.1  virginica
# 8           6.1         2.8          4.7         1.2 versicolor
# 9           4.6         3.4          1.4         0.3     setosa
# 10          6.3         2.5          4.9         1.5 versicolor

sample_frac(iris,0.1) #從iris里隨機(jī)抽取了10%（0.1）行的數(shù)據(jù)（iris數(shù)據(jù)框一共150行，返回了15行）
#    Sepal.Length Sepal.Width Petal.Length Petal.Width    Species
# 1           6.0         2.9          4.5         1.5 versicolor
# 2           5.5         3.5          1.3         0.2     setosa
# 3           6.5         3.0          5.8         2.2  virginica
# 4           7.2         3.6          6.1         2.5  virginica
# 5           5.5         4.2          1.4         0.2     setosa
# 6           7.6         3.0          6.6         2.1  virginica
# 7           7.2         3.2          6.0         1.8  virginica
# 8           5.6         3.0          4.1         1.3 versicolor
# 9           5.2         4.1          1.5         0.1     setosa
# 10          6.0         2.7          5.1         1.6 versicolor
# 11          5.6         2.5          3.9         1.1 versicolor
# 12          6.1         2.8          4.7         1.2 versicolor
# 13          4.5         2.3          1.3         0.3     setosa
# 14          6.5         3.2          5.1         2.0  virginica
# 15          5.1         3.8          1.5         0.3     setosa

7. group_by 分組函數(shù)??（可以根據(jù)數(shù)據(jù)框中的分類(lèi)變量進(jìn)行分組，然后結(jié)合summarise函數(shù)進(jìn)行匯總操作）

group=group_by(tips,smoker)
summarise(group,count=n(),mean_tips=mean(tip),sd_bill=sd(total_bill))
# A tibble: 2 x 4
#  smoker count mean_tips sd_bill
#  <fct>  <int>     <dbl>   <dbl>
# 1 No       151      2.99    8.26
# 2 Yes       93      3.01    9.83

使用group_by函數(shù)，根據(jù)smoker對(duì)tips進(jìn)行分組。之后采用summarize函數(shù)對(duì)分組數(shù)據(jù)進(jìn)行統(tǒng)計(jì)。如上分別計(jì)算了smoker和non-smoker的個(gè)數(shù)、均值和標(biāo)準(zhǔn)差

8. 管道符 %>%??

result <- tips %>% group_by(smoker,sex) %>% summarise(count = n(),mean_tips=mean(tip),sd_bill=sd(total_bill))
result
# A tibble: 4 x 5
# Groups:   smoker [2]
#  smoker sex    count mean_tips sd_bill
#  <fct>  <fct>  <int>     <dbl>   <dbl>
# 1 No     Female    54      2.77    7.29
# 2 No     Male      97      3.11    8.73
# 3 Yes    Female    33      2.93    9.19
# 4 Yes    Male      60      3.05    9.91

9. join函數(shù)家族（對(duì)數(shù)據(jù)框進(jìn)行合并）

• 9.1 inner_join函數(shù)（??和merge一樣）

 df_a <- data.frame(x=c('a','b','c','a','c','b','c'),y=1:7)
df_b <- data.frame(x=c('a','b','a'),z=10:12)
 inner_join(df_a,df_b,by='x')
#  x y  z
# 1 a 1 10
# 2 a 1 12
# 3 b 2 11
# 4 a 4 10
# 5 a 4 12
# 6 b 6 11

根據(jù)共有的x來(lái)對(duì)數(shù)據(jù)框進(jìn)行合并，由于第二個(gè)數(shù)據(jù)框中的x沒(méi)有c，因而c被刪掉了未被合并

• 9.2 semi_join函數(shù)

semi_join(df_a,df_b,by='x')
#   x y
# 1 a 1
# 2 b 2
# 3 a 4
# 4 b 6

與inner_join類(lèi)似，但只返回合并后的x和y

• 9.3 anti_join函數(shù)

anti_join(df_a,df_b,by='x')
#   x y
# 1 c 3
# 2 c 5
# 3 c 7

與semi_join完全相反，只返回兩個(gè)數(shù)據(jù)框中沒(méi)有重復(fù)的值

• 9.4 left_join

left_join(df_a,df_b,by='x')
#   x y  z
# 1 a 1 10
# 2 a 1 12
# 3 b 2 11
# 4 c 3 NA
# 5 a 4 10
# 6 a 4 12
# 7 c 5 NA
# 8 b 6 11
# 9 c 7 NA

兩個(gè)數(shù)據(jù)框合并時(shí)右邊的數(shù)據(jù)框向左邊的數(shù)據(jù)框合并，如果左邊的數(shù)據(jù)框有右邊數(shù)據(jù)框沒(méi)有的觀測(cè)，返回NA值。

• 9.5 right_join

right_join(df_a,df_b,by='x')
#   x y  z
# 1 a 1 10
# 2 a 1 12
# 3 b 2 11
# 4 a 4 10
# 5 a 4 12
# 6 b 6 11

兩個(gè)數(shù)據(jù)框合并時(shí)左邊的數(shù)據(jù)框向右邊的數(shù)據(jù)框合并，如果左邊的數(shù)據(jù)框有右邊數(shù)據(jù)框沒(méi)有的觀測(cè)，則不予顯示。

10. count函數(shù)（對(duì)list中針對(duì)某個(gè)分組變量的各個(gè)觀測(cè)值的數(shù)量進(jìn)行統(tǒng)計(jì)）??

count(tips,smoker)
#  smoker   n
#1     No 151
#2    Yes  93

11. summarise函數(shù)

summarise函數(shù)對(duì)數(shù)據(jù)進(jìn)行統(tǒng)計(jì)描述，可以將數(shù)據(jù)框折疊成一行，常與group_by函數(shù)搭配使用。group_by()與summarise()的組合構(gòu)成了使用dplyr包時(shí)最常用的操作之一：分組摘要。
比base包中的summary()更加靈活

mtcars %>%
  summarise(mean = mean(disp), n = n()) #查看disp這一列的均值，n = n()看有多少個(gè)觀測(cè)
#       mean  n
# 1 230.7219 32

# 根據(jù)某個(gè)變量對(duì)某一列分組并統(tǒng)計(jì)
mtcars %>%
  group_by(cyl) %>%
  summarise(mean = mean(disp), n = n())
#     cyl  mean     n
#   <dbl> <dbl> <int>
# 1     4  105.    11
# 2     6  183.     7
# 3     8  353.    14

# 同時(shí)進(jìn)行多種統(tǒng)計(jì)運(yùn)算
mtcars %>%
   group_by(cyl) %>%
   summarise(qs = quantile(disp, c(0.25, 0.75)), prob = c(0.25, 0.75))
# `summarise()` has grouped output by 'cyl'. You can override using the `.groups` argument.
#   A tibble: 6 x 3
#   Groups:   cyl [3]
#     cyl    qs  prob
#   <dbl> <dbl> <dbl>
# 1     4  78.8  0.25
# 2     4 121.   0.75
# 3     6 160    0.25
# 4     6 196.   0.75
# 5     8 302.   0.25
# 6     8 390    0.75

#更多應(yīng)用見(jiàn)?summarise

dplyr cheatsheet