一犬辰、數(shù)據(jù)的讀取與導(dǎo)出
①. 讀取數(shù)據(jù):txt嗦篱、csv和xlsx文件
· 速度最慢
options(stringsAsFactors=FALSE) ## 防止把字符串轉(zhuǎn)化成因子型變量
data <- read.table("iris.txt",sep="\t",header = T) ## 讀取txt文件
data <- read.csv("iris.csv",sep=",",header = T) ## 讀取csv文件
# install.packages("readxl");library(readxl)
data <- readxl::read_xlsx("iris.xlsx",sheet = 1) ## 讀取xlsx文件
· 中等速度
library(readr)
data <- read_tsv("iris.txt") ## 讀取txt文件
data <- read_csv("iris.csv") ## 讀取csv文件
data <- read_delim("iris.txt",delim = "\t") ## 讀取txt文件
data <- read_delim("iris.csv",delim = ",") ## 讀取csv文件
· 速度最快
library(data.table)
data <- data.table::fread("iris.txt") ## 讀取txt文件
data <- data.table::fread("iris.csv") ## 讀取csv文件
②. 數(shù)據(jù)導(dǎo)出：導(dǎo)出為txt、csv和xlsx文件
write.table(data ,"iris_new.txt",sep="\t",row.names = F) ## 導(dǎo)出txt文件
write.table(data ,"iris_new.csv",sep=",",row.names = F) ## 導(dǎo)出csv文件
write.csv(data ,"iris_new.csv") ## 導(dǎo)出csv文件
# install.packages("openxlsx");library(openxlsx)
write.xlsx(data ,"iris_new.xlsx",sheetName="iris") ## 導(dǎo)出xlsx文件
③. 環(huán)境變量的導(dǎo)出與讀取
saveRDS(data,"iris.rds")
save(data,file="iris.Rdata")
data <- readRDS("iris.rds")
二幌缝、dplyr清洗數(shù)據(jù)
· 本文章所使用的Iris數(shù)據(jù)集（鳶尾花數(shù)據(jù)集）灸促，是機(jī)器學(xué)習(xí)任務(wù)中常用的分類實(shí)驗(yàn)數(shù)據(jù)集，由Fisher在1936收集整理涵卵。
· Iris一共包含150個(gè)樣本浴栽，分為3類，每類50個(gè)數(shù)據(jù)轿偎，每個(gè)數(shù)據(jù)包含4個(gè)屬性典鸡。
· 可通過(guò)花萼長(zhǎng)度，花萼寬度坏晦，花瓣長(zhǎng)度萝玷，花瓣寬度4個(gè)特征預(yù)測(cè)鳶尾花卉屬于山鳶尾（Setosa）、變色鳶尾（Versicolour）還是維吉尼亞鳶尾（Virginica）中的哪一類昆婿。
> head(iris,10) # 查看數(shù)據(jù)集前10行
Sepal.Length Sepal.Width Petal.Length Petal.Width Species
1 5.1 3.5 1.4 0.2 setosa
2 4.9 3.0 1.4 0.2 setosa
3 4.7 3.2 1.3 0.2 setosa
4 4.6 3.1 1.5 0.2 setosa
5 5.0 3.6 1.4 0.2 setosa
6 5.4 3.9 1.7 0.4 setosa
7 4.6 3.4 1.4 0.3 setosa
8 5.0 3.4 1.5 0.2 setosa
9 4.4 2.9 1.4 0.2 setosa
10 4.9 3.1 1.5 0.1 setosa
install.packages("dplyr") # dplyr包安裝
· dplyr是一個(gè)很方便的用來(lái)清洗數(shù)據(jù)的R包球碉；
· 我們主要掌握它的6個(gè)函數(shù)，足以應(yīng)付大部分的需求：filter()仓蛆、arrange()睁冬、select()、mutate()看疙、summarize()和group_by()痴突；
①. filter()
· 篩選花萼長(zhǎng)度＞5搂蜓，寬度>3.5且物種為山鳶尾的行
> dplyr::filter(iris,Species=="setosa",Sepal.Length>5,Sepal.Width>3.5)
Sepal.Length Sepal.Width Petal.Length Petal.Width Species
1 5.4 3.9 1.7 0.4 setosa
2 5.4 3.7 1.5 0.2 setosa
3 5.8 4.0 1.2 0.2 setosa
4 5.7 4.4 1.5 0.4 setosa
5 5.4 3.9 1.3 0.4 setosa
6 5.7 3.8 1.7 0.3 setosa
7 5.1 3.8 1.5 0.3 setosa
8 5.1 3.7 1.5 0.4 setosa
9 5.2 4.1 1.5 0.1 setosa
10 5.5 4.2 1.4 0.2 setosa
11 5.1 3.8 1.9 0.4 setosa
12 5.1 3.8 1.6 0.2 setosa
13 5.3 3.7 1.5 0.2 setosa
· 篩選花萼長(zhǎng)度＞5且物種為山鳶尾，或花萼寬度>3.5且物種為山鳶尾的行
dplyr::filter(iris,Species=="setosa",Sepal.Length>5|Sepal.Width>3.5)
Sepal.Length Sepal.Width Petal.Length Petal.Width Species
1 5.1 3.5 1.4 0.2 setosa
2 5.0 3.6 1.4 0.2 setosa
3 5.4 3.9 1.7 0.4 setosa
4 5.4 3.7 1.5 0.2 setosa
5 5.8 4.0 1.2 0.2 setosa
6 5.7 4.4 1.5 0.4 setosa
7 5.4 3.9 1.3 0.4 setosa
8 5.1 3.5 1.4 0.3 setosa
9 5.7 3.8 1.7 0.3 setosa
10 5.1 3.8 1.5 0.3 setosa
11 5.4 3.4 1.7 0.2 setosa
12 5.1 3.7 1.5 0.4 setosa
13 4.6 3.6 1.0 0.2 setosa
14 5.1 3.3 1.7 0.5 setosa
15 5.2 3.5 1.5 0.2 setosa
16 5.2 3.4 1.4 0.2 setosa
17 5.4 3.4 1.5 0.4 setosa
18 5.2 4.1 1.5 0.1 setosa
19 5.5 4.2 1.4 0.2 setosa
20 5.5 3.5 1.3 0.2 setosa
21 4.9 3.6 1.4 0.1 setosa
22 5.1 3.4 1.5 0.2 setosa
23 5.1 3.8 1.9 0.4 setosa
24 5.1 3.8 1.6 0.2 setosa
25 5.3 3.7 1.5 0.2 setosa
②. arrange()
· 按列排序辽装，默認(rèn)是升序帮碰。我們可以通過(guò)參數(shù)desc()來(lái)使它降序
· 我們按花瓣長(zhǎng)度Petal.Length排序
> head(dplyr::arrange(iris,desc(Petal.Length)),5)
Sepal.Length Sepal.Width Petal.Length Petal.Width Species
1 7.7 2.6 6.9 2.3 virginica
2 7.7 3.8 6.7 2.2 virginica
3 7.7 2.8 6.7 2.0 virginica
4 7.6 3.0 6.6 2.1 virginica
5 7.9 3.8 6.4 2.0 virginica
> head(dplyr::arrange(iris,Petal.Length),5)
Sepal.Length Sepal.Width Petal.Length Petal.Width Species
1 4.6 3.6 1.0 0.2 setosa
2 4.3 3.0 1.1 0.1 setosa
3 5.8 4.0 1.2 0.2 setosa
4 5.0 3.2 1.2 0.2 setosa
5 4.7 3.2 1.3 0.2 setosa
③. select()
· 選擇我們需要的列，例如我們只希望保留花瓣長(zhǎng)度Petal.Length拾积、花瓣寬度 Petal.Width和物種Species的信息
> head(dplyr::select(iris,Petal.Length,Petal.Width,Species),5)
Petal.Length Petal.Width Species
1 1.4 0.2 setosa
2 1.4 0.2 setosa
3 1.3 0.2 setosa
4 1.5 0.2 setosa
5 1.4 0.2 setosa
· 剔除我們不需要的列殉挽，例如我們只希望保留花瓣長(zhǎng)度Petal.Length、花瓣寬度 Petal.Width和物種Species的信息拓巧，那我們可以剔除花萼長(zhǎng)度Sepal.Length和花萼寬度epal.Width
> head(dplyr::select(iris,-Sepal.Length,-Sepal.Width),5)
Petal.Length Petal.Width Species
1 1.4 0.2 setosa
2 1.4 0.2 setosa
3 1.3 0.2 setosa
4 1.5 0.2 setosa
5 1.4 0.2 setosa
④. mutate()
· 使用mutate()可以往數(shù)據(jù)集里面添加新的列
· 我們往iris數(shù)據(jù)集里面添加兩個(gè)新列斯碌，分別是花萼長(zhǎng)寬比Sepal.prop和花瓣長(zhǎng)寬比Petal..prop
>head(dplyr::mutate(iris,Sepal.prop=Sepal.Length/Sepal.Width,
Petal.prop=Petal.Length/Petal.Width),5)
Sepal.Length Sepal.Width Petal.Length Petal.Width Species Sepal.prop Petal.prop
1 5.1 3.5 1.4 0.2 setosa 1.457143 7.0
2 4.9 3.0 1.4 0.2 setosa 1.633333 7.0
3 4.7 3.2 1.3 0.2 setosa 1.468750 6.5
4 4.6 3.1 1.5 0.2 setosa 1.483871 7.5
5 5.0 3.6 1.4 0.2 setosa 1.388889 7.0
⑤. summarize()
· summarize()和mutate()一樣可以產(chǎn)生新變量，但summarize()不能保留原來(lái)的變量肛度；
> head(dplyr::summarize(iris,Sepal.prop=Sepal.Length/Sepal.Width,
+ Petal.prop=Petal.Length/Petal.Width),5)
Sepal.prop Petal.prop
1 1.457143 7.0
2 1.633333 7.0
3 1.468750 6.5
4 1.483871 7.5
5 1.388889 7.0
· summarize()更多的適用于摘要統(tǒng)計(jì)傻唾，即可以把許多行壓縮成一行。而mutate()無(wú)法做到這一點(diǎn)承耿；
>head(dplyr::summarize(iris,Sepal.Length.mean=mean(Sepal.Length)),5)
Sepal.Length.mean
1 5.843333
> head(dplyr::mutate(iris,Sepal.Length.mean=mean(Sepal.Length)),5)
Sepal.Length Sepal.Width Petal.Length Petal.Width Species Sepal.Length.mean
1 5.1 3.5 1.4 0.2 setosa 5.843333
2 4.9 3.0 1.4 0.2 setosa 5.843333
3 4.7 3.2 1.3 0.2 setosa 5.843333
4 4.6 3.1 1.5 0.2 setosa 5.843333
5 5.0 3.6 1.4 0.2 setosa 5.843333
⑥. group_by()
· group_by()可以按組操作
· group_by()只有和管道符(%>% )或者summarize()結(jié)合起來(lái)冠骄，才能達(dá)到最佳的效果
· 例如，計(jì)算不同物種花萼長(zhǎng)度的平均值
> iris %>% group_by(Species) %>% summarize(Sepal.Length.mean=mean(Sepal.Length))
# A tibble: 3 x 2
Species Sepal.Length.mean
* <fct> <dbl>
1 setosa 5.01
2 versicolor 5.94
3 virginica 6.59
· 同樣是計(jì)算不同物種花萼長(zhǎng)度的平均值加袋，mutate()無(wú)法將許多行壓縮成一行
> iris %>% group_by(Species) %>% mutate(Sepal.Length.mean=mean(Sepal.Length))
# A tibble: 150 x 6
# Groups: Species [3]
Sepal.Length Sepal.Width Petal.Length Petal.Width Species Sepal.Length.mean
<dbl> <dbl> <dbl> <dbl> <fct> <dbl>
1 5.1 3.5 1.4 0.2 setosa 5.01
2 4.9 3 1.4 0.2 setosa 5.01
3 4.7 3.2 1.3 0.2 setosa 5.01
4 4.6 3.1 1.5 0.2 setosa 5.01
5 5 3.6 1.4 0.2 setosa 5.01
6 5.4 3.9 1.7 0.4 setosa 5.01
7 4.6 3.4 1.4 0.3 setosa 5.01
8 5 3.4 1.5 0.2 setosa 5.01
9 4.4 2.9 1.4 0.2 setosa 5.01
10 4.9 3.1 1.5 0.1 setosa 5.01
# … with 140 more rows
⑦. 總結(jié)：
· 把這6個(gè)函數(shù)都用上
iris %>%
select(Sepal.Length,Petal.Length,Species) %>%
filter(Sepal.Length>1,Petal.Length>1) %>%
mutate(mean.Length=(Sepal.Length+Petal.Length)/2) %>%
group_by(Species) %>%
summarise(Length=mean(mean.Length))
# A tibble: 3 x 2
Species Length
* <fct> <dbl>
1 setosa 3.24
2 versicolor 5.10
3 virginica 6.07
三凛辣、表格操作
①. 表格的變換
· gather()函數(shù)可以把寬數(shù)據(jù)變成長(zhǎng)數(shù)據(jù)
> df <- data.frame(x=c("A","B","C"),'2010'=c(1,3,4),'2011'=c(3,5,2))
> colnames(df) <- c("name","2010","2011")
> df
name 2010 2011
1 A 1 3
2 B 3 5
3 C 4 2
> gather(df,year,value,-name)
name year value
1 A 2010 1
2 B 2010 3
3 C 2010 4
4 A 2011 3
5 B 2011 5
6 C 2011 2
> gather(df,year,value)
year value
1 name A
2 name B
3 name C
4 2010 1
5 2010 3
6 2010 4
7 2011 3
8 2011 5
9 2011 2
· spread()函數(shù)可以把長(zhǎng)數(shù)據(jù)變成寬數(shù)據(jù)
> df <- data.frame(x=c("A","B","C"),'2010'=c(1,3,4),'2011'=c(3,5,2))
> colnames(df) <- c("name","2010","2011")
> df <- gather(df,year,value,-name)
> df
name year value
1 A 2010 1
2 B 2010 3
3 C 2010 4
4 A 2011 3
5 B 2011 5
6 C 2011 2
> df <- spread(df,year,value)
> df
name 2010 2011
1 A 1 3
2 B 3 5
3 C 4 2
②. 表格的拼接
· cbind()函數(shù)縱向拼接
· rbind()函數(shù)縱向拼接
> df1 <- data.frame(x=c("a","b","c"),y=1:3)
> df2 <- data.frame(z=c("A","b","c"),y=c(3,6,9))
> cbind(df1,df2)
x y z y
1 a 1 A 3
2 b 2 b 6
3 c 3 c 9
> rbind(df1,df2)
Error in match.names(clabs, names(xi)) :
names do not match previous names
> colnames(df2)[1] <- "x"
> rbind(df1,df2)
x y
1 a 1
2 b 2
3 c 3
4 A 3
5 b 6
6 c 9
③. 表格的合并
· merge()函數(shù)
> df1 <- data.frame(x=c("a","b","c"),y=1:3)
> df2 <- data.frame(z=c("A","b","c"),y=c(3,6,9))
> merge(df1,df2,by.x = "x",by.y = "z")
x y.x y.y
1 b 2 6
2 c 3 9
> merge(df1,df2,by.x = "x",by.y = "z",all = T)
x y.x y.y
1 a 1 NA
2 A NA 3
3 b 2 6
4 c 3 9
· left_join函數(shù)、right_join函數(shù)职烧、inner_join函數(shù)和full_join函數(shù)
· left_join函數(shù)只保留左表的所有數(shù)據(jù)
> df1 <- data.frame(x=c("a","b","c"),y=1:3)
> df2 <- data.frame(z=c("A","b","c"),y=c(3,6,9))
> left_join(df1,df2)
Joining, by = "y"
x y z
1 a 1 <NA>
2 b 2 <NA>
3 c 3 A
> ?left_join
> left_join(df1,df2,by=c("x"="z"))
x y.x y.y
1 a 1 NA
2 b 2 6
3 c 3 9
· right_join函數(shù)只保留右表的所有數(shù)據(jù)
> df1 <- data.frame(x=c("a","b","c"),y=1:3)
> df2 <- data.frame(z=c("A","b","c"),y=c(3,6,9))
> right_join(df1,df2,by=c("x"="z"))
x y.x y.y
1 b 2 6
2 c 3 9
3 A NA 3
· inner_join函數(shù)只保留所有表的公共數(shù)據(jù)
> df1 <- data.frame(x=c("a","b","c"),y=1:3)
> df2 <- data.frame(z=c("A","b","c"),y=c(3,6,9))
> inner_join(df1,df2)
Joining, by = "y"
x y z
1 c 3 A
> inner_join(df1,df2,by=c("x"="z"))
x y.x y.y
1 b 2 6
2 c 3 9
· full_join函數(shù)保留所有表的公共數(shù)據(jù)
> df1 <- data.frame(x=c("a","b","c"),y=1:3)
> df2 <- data.frame(z=c("A","b","c"),y=c(3,6,9))
> full_join(df1,df2)
Joining, by = "y"
x y z
1 a 1 <NA>
2 b 2 <NA>
3 c 3 A
4 <NA> 6 b
5 <NA> 9 c
> full_join(df1,df2,by=c("x"="z"))
x y.x y.y
1 a 1 NA
2 b 2 6
3 c 3 9
4 A NA 3
