R语言大数据处理锦囊-data.table包

注，有任何建议或疑问,请加 QQ: 1595218767 ,共同探讨学习
如R/python代码编程作图等方面需要帮忙，欢迎来店咨询 之恒科技, 挥动热情的小爪爪期待你哦

前几天网友遇到一简单的freq统计,由于元数据有 18000000 行,运行半天没出结果,apply/分布式 ...方法尝试了多种, 遂推荐data.table给他, 效果豁然

m1 <- data.table::data.table(m2=sample(1:10000,18000000,replace = TRUE))
system.time(m1[,.N,m2])
用户 系统 流逝 
1.60 0.03 1.64 
system.time(table(m1$m2))
 用户  系统  流逝 
14.63  0.25 14.94

A. data.table简介

1. 基本架构

syntax1

Note: what to do means select/summarise/update/mutate...

syntax2

2. 特殊符号定义函数

Functions What for
.N n(),nrow(dt),length(dt$col)
.SD all cols except ones for by/keyby
.SDcols cols for .SD
.I rowIndex; seq_len(nrow(dt))
.GRP groupid of each group; seq(1,length(distinct(group)))
.BY merge_by, 用法见 数据集合并

B. 使用笔记

1. 数据读取和存取

1.1 读取 (data)

1.1.1 fread

fread('path',header=TRUE，encoding='UTF-8')

dt4 <- fread('F:/360Downloads/aa.csv',header = TRUE)
str(dt4)
Classes ‘data.table’ and 'data.frame':  26 obs. of  3 variables:
 $ x: int  1 1 2 2 3 3 4 4 5 5 ...
 $ y: chr  "IR_consensus_ALT" "self_consensus_ALT" "IR_consensus_ALT" "self_consensus_ALT" ...
 $ z: int  20 29 26 14 10 7 399 346 439 28 ...
 - attr(*, ".internal.selfref")=<externalptr>
1.1.2 其他类型转化

data.table(data.frame/matrix) .. do.call(cbind,lapply(ls1,function(x)do.call(cbind,x)))

# --- from data.frame to data.table
dt2 <- data.table(iris) 
str(dt2)
Classes ‘data.table’ and 'data.frame':  150 obs. of  5 variables:
 $ Sepal.Length: num  5.1 4.9 4.7 4.6 5 5.4 4.6 5 4.4 4.9 ...
 $ Sepal.Width : num  3.5 3 3.2 3.1 3.6 3.9 3.4 3.4 2.9 3.1 ...
 $ Petal.Length: num  1.4 1.4 1.3 1.5 1.4 1.7 1.4 1.5 1.4 1.5 ...
 $ Petal.Width : num  0.2 0.2 0.2 0.2 0.2 0.4 0.3 0.2 0.2 0.1 ...
 $ Species     : Factor w/ 3 levels "setosa","versicolor",..: 1 1 1 1 1 1 1 1 1 1 ...
 - attr(*, ".internal.selfref")=<externalptr> 
# --- from matrix to data.table
(mx1 <- matrix(1:12,ncol=4))
     [,1] [,2] [,3] [,4]
[1,]    1    4    7   10
[2,]    2    5    8   11
[3,]    3    6    9   12
> (data.table(mx1 <- matrix(1:12,ncol=4)))
   V1 V2 V3 V4
1:  1  4  7 10
2:  2  5  8 11
3:  3  6  9 12
# --- from list to data.table (最底层list的规格必须一致)
ls1 <- list(a1=list(a11=c(1:3),a12=letters[4:6]),b1=list(b11=rep('nokia',3),b12=rep('good',3)))
str(ls1)
List of 2
 $ a1:List of 2
  ..$ a11: int [1:3] 1 2 3
  ..$ a12: chr [1:3] "d" "e" "f"
 $ b1:List of 2
  ..$ b11: chr [1:3] "nokia" "nokia" "nokia"
  ..$ b12: chr [1:3] "good" "good" "good"
(ls2 <- lapply(ls1,function(x)do.call(cbind,x)))
$a1
     a11 a12
[1,] "1" "d"
[2,] "2" "e"
[3,] "3" "f"
$b1
     b11     b12   
[1,] "nokia" "good"
[2,] "nokia" "good"
[3,] "nokia" "good"
do.call(cbind,ls2)
     a11 a12 b11     b12   
[1,] "1" "d" "nokia" "good"
[2,] "2" "e" "nokia" "good"
[3,] "3" "f" "nokia" "good"
1.1.3 直接生成 data.table

data.table(var1=c(),var2=c())

dt3 <- data.table(a1=c(1:6),a2=rep(1:3,2))
str(dt3)
Classes ‘data.table’ and 'data.frame':  6 obs. of  2 variables:
 $ a1: int  1 2 3 4 5 6
 $ a2: int  1 2 3 1 2 3
 - attr(*, ".internal.selfref")=<externalptr>

1.2 存取 (save)

fwrite(dt1,'path/name1.type',row.names=FALSE)

fwrite(x, file = "",na = "", dec = ".", row.names = FALSE, col.names = TRUE)
fwrite(dt1,'./dt1.csv',row.names=FALSE)

与 write.csv 等函数相较, fwrite 缺少 fileEncoding 参数, 参数(fileEncoding='UTF-8')会对windows 平台对有很大帮助,但该包的maintainer截止现在还没有将其加进维护list的打算...如果你也需要这个参数,请 需求+1 in fwrite.utf-8

2. 数据排序 (arrange)

dt[order(var1,-var2),] .. setkey(dt,var1,var2)

dt3 <- data.table(a1=c(1:6),a2=rep(1:3,2))
# --- order
dt3[order(a2,-a1)]
   a1 a2
1:  4  1
2:  1  1
3:  5  2
4:  2  2
5:  6  3
6:  3  3
dt3[,.SD,keyby=a2]
   a2 a1
1:  1  1
2:  1  4
3:  2  2
4:  2  5
5:  3  3
6:  3  6
# --- setkey
setkey(dt3,a2)
dt3
   a1 a2
1:  1  1
2:  4  1
3:  2  2
4:  5  2
5:  3  3
6:  6  3

3. 数据分组和统计 (group_by,summarise)

dt[,func(x),by/keyby=.(var1,var2)]

set.seed(123)
m1 <- data.table::data.table(m2=sample(1:10000,18000000,replace = TRUE))
m3 <- m1[,.N,m2]
head(m3)
     m2   N
1:  220 846
2: 6271 816
3: 8731 797
4: 5147 813
5: 6910 801
6: 6690 775
m3 <- m1[,.N,keyby=m2]
head(m3)
   m2   N
1:  1 782
2:  2 785
3:  3 834
4:  4 769
5:  5 808
6:  6 785
m4 <- m1[order(m2),.(cnt1=.N,mean1=mean(m2)),m2]
head(m4)
   m2 cnt1 mean1
1:  1  782     1
2:  2  785     2
3:  3  834     3
4:  4  769     4
5:  5  808     5
6:  6  785     6
dim(m4)
[1] 10000     3
dim(m3)
[1] 10000     2
dim(m1)
[1] 18000000       2
# --- special use of .N
m1[.N] # last row
      m2 cnt1
1: 10000  851
# --- .SD
dt4 <- data.table(a1=1:10,a2=rep(c(1,2),each=5),a3=rep(c(3,4),5),a4=runif(10,100,1000))
    a1 a2 a3       a4
 1:  1  1  3 804.9962
 2:  2  1  4 403.3980
 3:  3  1  3 212.7495
 4:  4  1  4 872.7093
 5:  5  1  3 744.6552
 6:  6  2  4 760.6584
 7:  7  2  3 227.0887
 8:  8  2  4 643.6969
 9:  9  2  3 489.9786
10: 10  2  4 458.0811
dt4[,lapply(.SD,mean),by=.(a2,a3)] # apply mean to all cols
   a2 a3 a1       a4
1:  1  3  3 587.4670
2:  1  4  3 638.0536
3:  2  4  8 620.8121
4:  2  3  8 358.5337
dt4[,lapply(.SD,mean),by=.(a2,a3),.SDcols=c('a4')] # apply mean to all cols of .SDcols
   a2 a3       a4
1:  1  3 587.4670
2:  1  4 638.0536
3:  2  4 620.8121
4:  2  3 358.5337
dt4[,lapply(.(a1,a4),mean),by=.(a2,a3)]
   a2 a3 V1       V2
1:  1  3  3 587.4670
2:  1  4  3 638.0536
3:  2  4  8 620.8121
4:  2  3  8 358.5337
# --- c() and .() in what-to-do 
dt4[,c(.N,lapply(.(a1,a4),mean)),by=.(a2,a3)]
   a2 a3 V1 V2       V3
1:  1  3  3  3 587.4670
2:  1  4  2  3 638.0536
3:  2  4  3  8 620.8121
4:  2  3  2  8 358.5337
dt4[,.(.N,lapply(.(a1,a4),mean)),by=.(a2,a3)]
   a2 a3 N       V2
1:  1  3 3        3
2:  1  3 3  587.467
3:  1  4 2        3
4:  1  4 2 638.0536
5:  2  4 3        8
6:  2  4 3 620.8121
7:  2  3 2        8
8:  2  3 2 358.5337
# --- topN in each group
dt4[,head(.SD,2),by=.(a2)] # top2 in each group
   a2 a1 a3       a4
1:  1  1  3 804.9962
2:  1  2  4 403.3980
3:  2  6  4 760.6584
4:  2  7  3 227.0887
dt4[,.SD[1:2],by=.(a2)] # top2 in each group
   a2 a1 a3       a4
1:  1  1  3 804.9962
2:  1  2  4 403.3980
3:  2  6  4 760.6584
4:  2  7  3 227.0887
dt4[order(a3),head(.SD,2),by=.(a2)] # top2 in each group ordered by cols
   a2 a1 a3       a4
1:  1  1  3 804.9962
2:  1  3  3 212.7495
3:  2  7  3 227.0887
4:  2  9  3 489.9786
# --- .I
dt4[, .(indx1=.I[which.max(a4)],val1=max(a4)), by=a2] # max(a4) index and value
   a2 indx1     val1
1:  1     4 872.7093
2:  2     6 760.6584
dt4[,.I[1:2],keyby=.(a2)] # row num of top2 in each group
   a2 V1
1:  1  1
2:  1  2
3:  2  6
4:  2  7
# --- .GRP
dt4[,.(ids1=.GRP),by=a3] # Groupid of group(seq(1,n_distinct(group)))
   a3 ids1
1:  3    1
2:  4    2

4. 数据筛选

4.1 行筛选 (filter)

dt[rowIndex,] .. dt[conditions,]

iris3 <- data.table(iris)
iris3[1:3,]
   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
iris3[1:3]
   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
iris3[c(1,3)]
   Sepal.Length Sepal.Width Petal.Length Petal.Width Species
1:          5.1         3.5          1.4         0.2  setosa
2:          4.7         3.2          1.3         0.2  setosa
iris3[Sepal.Length>=7&Petal.Width>=2.3]
   Sepal.Length Sepal.Width Petal.Length Petal.Width   Species
1:          7.2         3.6          6.1         2.5 virginica
2:          7.7         2.6          6.9         2.3 virginica
3:          7.7         3.0          6.1         2.3 virginica
iris3[Sepal.Length>=6&Petal.Width>=2][,c(2,4,5),with=FALSE][1:3]
   Sepal.Width Petal.Width   Species
1:         3.3         2.5 virginica
2:         3.0         2.1 virginica
3:         3.0         2.2 virginica
iris3[Sepal.Length>=6][Petal.Width>=2][,c(2,4,5),with=FALSE][1:3]
   Sepal.Width Petal.Width   Species
1:         3.3         2.5 virginica
2:         3.0         2.1 virginica
3:         3.0         2.2 virginica

4.2 列筛选 (select)

dt[,.(var1,var2)] .. dt[,list(var1,var2)]

iris3[1:3,.(Sepal.Length,Petal.Width,Species)]
   Sepal.Length Petal.Width Species
1:          5.1         0.2  setosa
2:          4.9         0.2  setosa
3:          4.7         0.2  setosa
iris3[1:3,1:4]
   Sepal.Length Sepal.Width Petal.Length Petal.Width
1:          5.1         3.5          1.4         0.2
2:          4.9         3.0          1.4         0.2
3:          4.7         3.2          1.3         0.2
iris3[1:3,c(1,3,4),with=FALSE]
   Sepal.Length Petal.Length Petal.Width
1:          5.1          1.4         0.2
2:          4.9          1.4         0.2
3:          4.7          1.3         0.2
iris3[1:3,c(-5),with=FALSE]
   Sepal.Length Sepal.Width Petal.Length Petal.Width
1:          5.1         3.5          1.4         0.2
2:          4.9         3.0          1.4         0.2
3:          4.7         3.2          1.3         0.2
iris3[c(1,3,5),-('Species'),with=FALSE]
   Sepal.Length Sepal.Width Petal.Length Petal.Width
1:          5.1         3.5          1.4         0.2
2:          4.7         3.2          1.3         0.2
3:          5.0         3.6          1.4         0.2

5. 生成新变量 或者 变量值修改

dt[,newVar:=func(var1)] .. dt[condition,Var1:='value']

> (dt3 <- data.table(a1=c(1:6),a2=rep(1:3,2),a3=c(letters[c(1,2,2,3,5)],NA)))
   a1 a2 a3
1:  1  1  a
2:  2  2  b
3:  3  3  b
4:  4  1  c
5:  5  2  e
6:  6  3 NA
> dt3[,a4:=a1+a2]
> dt3
   a1 a2 a3 a4
1:  1  1  a  2
2:  2  2  b  4
3:  3  3  b  6
4:  4  1  c  5
5:  5  2  e  7
6:  6  3 NA  9
> dt3[a2>2,a4:=NA]
> dt3
   a1 a2 a3 a4
1:  1  1  a  2
2:  2  2  b  4
3:  3  3  b NA
4:  4  1  c  5
5:  5  2  e  7
6:  6  3 NA NA

6. 数据拆分 (split)

split(dt,by=c('var1','var2'))

dt4 <- data.table(a1=1:10,a2=rep(c(1,2),each=5),a3=rep(c(3,4),5),a4=runif(10,100,1000))
> spl1 <- split(dt4,by=c('a2','a3'))
> spl2 <- split(dt4,by=c('a2','a3'),flatten = FALSE)
> str(spl1)
List of 4
 $ 1.3:Classes ‘data.table’ and 'data.frame':   3 obs. of  4 variables:
  ..$ a1: int [1:3] 1 3 5
  ..$ a2: num [1:3] 1 1 1
  ..$ a3: num [1:3] 3 3 3
  ..$ a4: num [1:3] 339 141 819
  ..- attr(*, ".internal.selfref")=<externalptr> 
 $ 1.4:Classes ‘data.table’ and 'data.frame':   2 obs. of  4 variables:
  ..$ a1: int [1:2] 2 4
  ..$ a2: num [1:2] 1 1
  ..$ a3: num [1:2] 4 4
  ..$ a4: num [1:2] 872 498
  ..- attr(*, ".internal.selfref")=<externalptr> 
 $ 2.4:Classes ‘data.table’ and 'data.frame':   3 obs. of  4 variables:
  ..$ a1: int [1:3] 6 8 10
  ..$ a2: num [1:3] 2 2 2
  ..$ a3: num [1:3] 4 4 4
  ..$ a4: num [1:3] 210 286 778
  ..- attr(*, ".internal.selfref")=<externalptr> 
 $ 2.3:Classes ‘data.table’ and 'data.frame':   2 obs. of  4 variables:
  ..$ a1: int [1:2] 7 9
  ..$ a2: num [1:2] 2 2
  ..$ a3: num [1:2] 3 3
  ..$ a4: num [1:2] 605 215
  ..- attr(*, ".internal.selfref")=<externalptr>
> str(spl2)
List of 2
 $ 1:List of 2
  ..$ 3:Classes ‘data.table’ and 'data.frame':  3 obs. of  4 variables:
  .. ..$ a1: int [1:3] 1 3 5
  .. ..$ a2: num [1:3] 1 1 1
  .. ..$ a3: num [1:3] 3 3 3
  .. ..$ a4: num [1:3] 339 141 819
  .. ..- attr(*, ".internal.selfref")=<externalptr> 
  ..$ 4:Classes ‘data.table’ and 'data.frame':  2 obs. of  4 variables:
  .. ..$ a1: int [1:2] 2 4
  .. ..$ a2: num [1:2] 1 1
  .. ..$ a3: num [1:2] 4 4
  .. ..$ a4: num [1:2] 872 498
  .. ..- attr(*, ".internal.selfref")=<externalptr> 
 $ 2:List of 2
  ..$ 4:Classes ‘data.table’ and 'data.frame':  3 obs. of  4 variables:
  .. ..$ a1: int [1:3] 6 8 10
  .. ..$ a2: num [1:3] 2 2 2
  .. ..$ a3: num [1:3] 4 4 4
  .. ..$ a4: num [1:3] 210 286 778
  .. ..- attr(*, ".internal.selfref")=<externalptr> 
  ..$ 3:Classes ‘data.table’ and 'data.frame':  2 obs. of  4 variables:
  .. ..$ a1: int [1:2] 7 9
  .. ..$ a2: num [1:2] 2 2
  .. ..$ a3: num [1:2] 3 3
  .. ..$ a4: num [1:2] 605 215
  .. ..- attr(*, ".internal.selfref")=<externalptr> 
> spl1$`1.3`
   a1 a2 a3       a4
1:  1  1  3 339.3754
2:  3  1  3 141.2481
3:  5  1  3 819.0324
> spl2$`1`
$`3`
   a1 a2 a3       a4
1:  1  1  3 339.3754
2:  3  1  3 141.2481
3:  5  1  3 819.0324

$`4`
   a1 a2 a3       a4
1:  2  1  4 872.0449
2:  4  1  4 497.9801

> spl2$`1`$`3`
   a1 a2 a3       a4
1:  1  1  3 339.3754
2:  3  1  3 141.2481
3:  5  1  3 819.0324

# --- 利用split分组实现组内标准化
library(openxlsx)
library(data.table)
call1<-read.xlsx("./360Downloads/data.xlsx",sheet=1)
data1 <- data.table(call1)[,lapply(.SD,sum),keyby=.(wangdian,weekon),.SDcols=names(call)[4:17]]
head(data1[,1:9],10)
    wangdian weekon value1 value2 value3 value4 value5 value6 value7
 1:   010ADA      1      0      0      2      2      0      0      0
 2:   010ADA      2      0      0      6      3      2      0      1
 3:   010ADA      3      0      0      1      0      1      0      0
 4:   010ADA      4      0      0      6      3      3      0      0
 5:   010ADA      5      0      0      8      3      5      0      0
 6:   010AFA      1      0      0      9      2      7      0      0
 7:   010AFA      2      0      0     21      2     19      0      0
 8:   010AFA      3      0      0     21      4     17      0      0
 9:   010AFA      4      0      0     31      5     26      0      0
10:   010AFA      5      0      0     37      9     28      0      0
spl1 <- split(data1,by=c('wangdian'))
res1 <- cbind(data1[,1:2],do.call(rbind,lapply(spl1,function(x) do.call(cbind,lapply(x[,3:16],scale)))))
> head(res1[,1:9],10)
    wangdian weekon  V1  V2         V3         V4         V5  V6         V7
 1:   010ADA      1 NaN NaN -0.8764598 -0.1533930 -1.1437255 NaN -0.4472136
 2:   010ADA      2 NaN NaN  0.4719399  0.6135720 -0.1039750 NaN  1.7888544
 3:   010ADA      3 NaN NaN -1.2135598 -1.6873230 -0.6238503 NaN -0.4472136
 4:   010ADA      4 NaN NaN  0.4719399  0.6135720  0.4159002 NaN -0.4472136
 5:   010ADA      5 NaN NaN  1.1461398  0.6135720  1.4556507 NaN -0.4472136
 6:   010AFA      1 NaN NaN -1.3789086 -0.8330522 -1.4895499 NaN        NaN
 7:   010AFA      2 NaN NaN -0.2608746 -0.8330522 -0.0480500 NaN        NaN
 8:   010AFA      3 NaN NaN -0.2608746 -0.1388420 -0.2883000 NaN        NaN
 9:   010AFA      4 NaN NaN  0.6708204  0.2082630  0.7928250 NaN        NaN
10:   010AFA      5 NaN NaN  1.2298374  1.5966833  1.0330750 NaN        NaN

7. 数据集合并 (join)

merge(dt1,dt2,by.x='var1',by.y='var2',all.x=FALSE,all.y=FALSE,all=TRUE)
setkey(dt,var)..dt1[dt2[on=c('var1'='var2')]]

> dt5 <- data.table(a1=1:5,a2=letters[c(1,2,3,5,6)])
> dt6 <- data.table(a3=6:10,a4=letters[c(2,3,3,6,7)])
> dt5
   a1 a2
1:  1  a
2:  2  b
3:  3  c
4:  4  e
5:  5  f
> dt6
   a3 a4
1:  6  b
2:  7  c
3:  8  c
4:  9  f
5: 10  g
> merge(dt5,dt6,by.x='a2',by.y='a4') # inner jion
   a2 a1 a3
1:  b  2  6
2:  c  3  7
3:  c  3  8
4:  f  5  9
> merge(dt5,dt6,by.x='a2',by.y='a4',all = TRUE) # full join
   a2 a1 a3
1:  a  1 NA
2:  b  2  6
3:  c  3  7
4:  c  3  8
5:  e  4 NA
6:  f  5  9
7:  g NA 10
> merge(dt5,dt6,by.x='a2',by.y='a4',all.x = TRUE) # left join
   a2 a1 a3
1:  a  1 NA
2:  b  2  6
3:  c  3  7
4:  c  3  8
5:  e  4 NA
6:  f  5  9
> merge(dt5,dt6,by.x='a2',by.y='a4',all.y = TRUE) # right join
   a2 a1 a3
1:  b  2  6
2:  c  3  7
3:  c  3  8
4:  f  5  9
5:  g NA 10
> dt5[dt5$a2 %in% dt6$a4] # semi join
   a1 a2
1:  2  b
2:  3  c
3:  5  f
> dt5[!(dt5$a2 %in% dt6$a4)] # anti join
   a1 a2
1:  1  a
2:  4  e
# --- setkey
> setkey(dt5,a2)
> setkey(dt6,a4)
> tables()
> dt5[dt6,on=c('a2'='a4')] # right join
   a1 a2 a3
1:  2  b  6
2:  3  c  7
3:  3  c  8
4:  5  f  9
5: NA  g 10
> dt6[dt5,on=c('a4'='a2')] # left join
   a3 a4 a1
1: NA  a  1
2:  6  b  2
3:  7  c  3
4:  8  c  3
5: NA  e  4
6:  9  f  5

8. 数据变形(melt/dcast)

melt(dt,id.vars=c('var1','var2'),measure.vars=c('var3','var4'),variable.name='',value.name='')
dcast(dt,formula(rows~cols),value.Var='',fun.aggregate=func)

> (dt3 <- data.table(a1=c(1:6),a2=rep(1:3,2),a3=c(letters[c(1,2,2,3,5)],NA)))
   a1 a2 a3
1:  1  1  a
2:  2  2  b
3:  3  3  b
4:  4  1  c
5:  5  2  e
6:  6  3 NA
> melt(dt3,id.vars = c('a1'),measure.vars = c('a2'))
   a1 variable value
1:  1       a2     1
2:  2       a2     2
3:  3       a2     3
4:  4       a2     1
5:  5       a2     2
6:  6       a2     3
> (dt4 <- melt(dt3,id.vars = c('a1'),measure.vars = c('a2','a3'),variable.name = 'meltVar',value.name = 'value1'))
    a1 meltVar value1
 1:  1      a2      1
 2:  2      a2      2
 3:  3      a2      3
 4:  4      a2      1
 5:  5      a2      2
 6:  6      a2      3
 7:  1      a3      a
 8:  2      a3      b
 9:  3      a3      b
10:  4      a3      c
11:  5      a3      e
12:  6      a3     NA
Warning message:
In melt.data.table(dt3, id.vars = c("a1"), measure.vars = c("a2",  :
  'measure.vars' [a2, a3] are not all of the same type. By order of hierarchy, the molten data value column will be of type 'character'. All measure variables not of type 'character' will be coerced to. Check DETAILS in ?melt.data.table for more on coercion.
> dt4[,lapply(.SD,class)]
        a1 variable     value
1: integer   factor character
# --- dcast
> dcast(dt4,a1~meltVar,value.var = 'value1')
   a1 a2 a3
1:  1  1  a
2:  2  2  b
3:  3  3  b
4:  4  1  c
5:  5  2  e
6:  6  3 NA
> d1 <- data.table(a1=rep(1:2,each=3),a2=rep(1:2,3),a3=1:6)
> dcast(d1,a1~a2,value.var = 'a3',fun.aggregate = sum)
   a1 1  2
1:  1 4  2
2:  2 5 10
> dcast(d1,a1~a2,value.var = 'a3')
Aggregate function missing, defaulting to 'length'
   a1 1 2
1:  1 2 1
2:  2 1 2

9. 列名称修改

setnames(dt,c(index1,index2),c('name1','names2'))
setnames(dt,c('name3','name4'),c('name1','names2'))

dt3 <- data.table(a1=c(1:6),a2=rep(1:3,2))
> names(dt3)
[1] "a1" "a2"
> setnames(dt3,c('a1','a2'),c('a3','a4'))
> names(dt3)
[1] "a3" "a4"
> setnames(dt3,c(1,2),c('a5','a6'))
> names(dt3)
[1] "a5" "a6"

.SD, .SDcols 特辑

library(openxlsx)
library(data.table)
call<-read.xlsx("G:/rwork/wuliu/distribution1.xlsx",sheet=1)
# --- 按照wangdian, weekon分组对剩余变量求和
data1 <- data.table(call)[,lapply(.SD,sum),
                          keyby=.(wangdian,weekon),
                          .SDcols=names(call)[4:17]]
head(data1)

   wangdian weekon linshengzhida kuashengjian G-G
1:   010ADA      1             0            2   2
2:   010ADA      2             0            6   3
3:   010ADA      3             0            1   0
4:   010ADA      4             0            6   3
5:   010ADA      5             0            8   3
6:   010AFA      1             0            9   2

# --- 按照 wangdian 分组将各week的变量值完成标准化处理(scale)
spl1 <- split(data1,by=c('wangdian'))
res1 <- do.call(rbind,lapply(spl1,function(x) do.call(cbind,lapply(x[,3:16],scale))))
res2 <- cbind(data1[,1:2],res1)
colnames(res2) <- colnames(data1)
head(res2)

    wangdian weekon linshengzhida kuashengjian        G-G
 1:   010ADA      1           NaN   -0.8764598 -0.1533930
 2:   010ADA      2           NaN    0.4719399  0.6135720
 3:   010ADA      3           NaN   -1.2135598 -1.6873230
 4:   010ADA      4           NaN    0.4719399  0.6135720
 5:   010ADA      5           NaN    1.1461398  0.6135720
 6:   010AFA      1           NaN   -1.3789086 -0.8330522

# --- NA 替换成 0
res3 <- res2[,c(.(wangdian=wangdian,weekon=weekon),
                lapply(.SD,function(x) ifelse(is.na(x),0,x))),
             .SDcols=colnames(res2)[-c(1:2)]]
head(res3)

    wangdian weekon linshengzhida kuashengjian        G-G
 1:   010ADA      1             0   -0.8764598 -0.1533930
 2:   010ADA      2             0    0.4719399  0.6135720
 3:   010ADA      3             0   -1.2135598 -1.6873230
 4:   010ADA      4             0    0.4719399  0.6135720
 5:   010ADA      5             0    1.1461398  0.6135720
 6:   010AFA      1             0   -1.3789086 -0.8330522

参考资料

Rdocument of data.table
Rdatatable github wiki

最后编辑于 ：
©著作权归作者所有,转载或内容合作请联系作者
  • 人面猴
    序言：七十年代末，一起剥皮案震惊了整个滨河市，随后出现的几起案子，更是在滨河造成了极大的恐慌，老刑警刘岩，带你破解...
    沈念sama阅读 211,194评论 6赞 490
  • 死咒
    序言：滨河连续发生了三起死亡事件，死亡现场离奇诡异，居然都是意外死亡，警方通过查阅死者的电脑和手机，发现死者居然都...
    沈念sama阅读 90,058评论 2赞 385
  • 救了他两次的神仙让他今天三更去死
    文/潘晓璐 我一进店门，熙熙楼的掌柜王于贵愁眉苦脸地迎上来，“玉大人，你说我怎么就摊上这事。” “怎么了？”我有些...
    开封第一讲书人阅读 156,780评论 0赞 346
  • 道士缉凶录：失踪的卖姜人
    文/不坏的土叔 我叫张陵，是天一观的道长。 经常有香客问我，道长，这世上最难降的妖魔是什么？ 我笑而不...
    开封第一讲书人阅读 56,388评论 1赞 283
  • 港岛之恋（遗憾婚礼）
    正文 为了忘掉前任，我火速办了婚礼，结果婚礼上，老公的妹妹穿的比我还像新娘。我一直安慰自己，他们只是感情好，可当我...
    茶点故事阅读 65,430评论 5赞 384
  • 恶毒庶女顶嫁案：这布局不是一般人想出来的
    文/花漫 我一把揭开白布。 她就那样静静地躺着，像睡着了一般。 火红的嫁衣衬着肌肤如雪。 梳的纹丝不乱的头发上，一...
    开封第一讲书人阅读 49,764评论 1赞 290
  • 城市分裂传说
    那天，我揣着相机与录音，去河边找鬼。 笑死，一个胖子当着我的面吹牛，可吹牛的内容都是我干的。 我是一名探鬼主播，决...
    沈念sama阅读 38,907评论 3赞 406
  • 双鸳鸯连环套：你想象不到人心有多黑
    文/苍兰香墨 我猛地睁开眼，长吁一口气：“原来是场噩梦啊……” “哼！你这毒妇竟也来了？” 一声冷哼从身侧响起，我...
    开封第一讲书人阅读 37,679评论 0赞 266
  • 万荣杀人案实录
    序言：老挝万荣一对情侣失踪，失踪者是张志新（化名）和其女友刘颖，没想到半个月后，有当地人在树林里发现了一具尸体，经...
    沈念sama阅读 44,122评论 1赞 303
  • 护林员之死
    正文 独居荒郊野岭守林人离奇死亡，尸身上长有42处带血的脓包…… 初始之章·张勋 以下内容为张勋视角 年9月15日...
    茶点故事阅读 36,459评论 2赞 325
  • 白月光启示录
    正文 我和宋清朗相恋三年，在试婚纱的时候发现自己被绿了。 大学时的朋友给我发了我未婚夫和他白月光在一起吃饭的照片。...
    茶点故事阅读 38,605评论 1赞 340
  • 活死人
    序言：一个原本活蹦乱跳的男人离奇死亡，死状恐怖，灵堂内的尸体忽然破棺而出，到底是诈尸还是另有隐情，我是刑警宁泽，带...
    沈念sama阅读 34,270评论 4赞 329
  • 日本核电站爆炸内幕
    正文 年R本政府宣布，位于F岛的核电站，受9级特大地震影响，放射性物质发生泄漏。R本人自食恶果不足惜，却给世界环境...
    茶点故事阅读 39,867评论 3赞 312
  • 男人毒药：我在死后第九天来索命
    文/蒙蒙 一、第九天 我趴在偏房一处隐蔽的房顶上张望。 院中可真热闹，春花似锦、人声如沸。这庄子的主人今日做“春日...
    开封第一讲书人阅读 30,734评论 0赞 21
  • 一桩弑父案，背后竟有这般阴谋
    文/苍兰香墨 我抬头看了看天上的太阳。三九已至，却和暖如春，着一层夹袄步出监牢的瞬间，已是汗流浃背。 一阵脚步声响...
    开封第一讲书人阅读 31,961评论 1赞 265
  • 情欲美人皮
    我被黑心中介骗来泰国打工， 没想到刚下飞机就差点儿被人妖公主榨干…… 1. 我叫王不留，地道东北人。 一个月前我还...
    沈念sama阅读 46,297评论 2赞 360
  • 代替公主和亲
    正文 我出身青楼，却偏偏与公主长得像，于是被迫代替她去往敌国和亲。 传闻我的和亲对象是个残疾皇子，可洞房花烛夜当晚...
    茶点故事阅读 43,472评论 2赞 348

推荐阅读更多精彩内容