在R中选择数据表中的列
我有如下所示的在R中选择数据表中的列,r,data.table,R,Data.table,我有如下所示的“in_table”。我需要使用“Comb\u table”获取“Table1”、“Table2”、“Table3”等。基本上，当Comb_表中的变量为1时，我需要将其包含在列表中 有没有什么有效的方法可以用R语言而不是手动输入所有的组合 感谢您的帮助 谢谢 in_table: POL Var1 Var2 Var3 Var4 Var5 Var6 Var7 8035 1 11 1 GRD 0030 0110 09
“in_table”“Comb\u table”“Table1”、“Table2”、“Table3”等。基本上，当Comb_表中的变量为1时，我需要将其包含在列表中
有没有什么有效的方法可以用R语言而不是手动输入所有的组合
感谢您的帮助
谢谢
in_table:
POL Var1 Var2 Var3 Var4 Var5 Var6 Var7
8035 1 11 1 GRD 0030 0110 09/30
36763 1 88 13 GRD 5260 0300 11/15
36763 1 88 13 GRD 5280 0300 11/15
35786 1 88 13 GRD 0030 0110 09/30
Comb_table:
Var1 Var2 Var3 Var4 Var5 Var6 Var7
Table1 1 1 1 1 1 1 1
Table2 0 1 1 1 1 1 1
Table3 1 0 1 1 1 1 1
Table1 <- in_table[, .(Pol_count = length(unique(POL))), by = list(Var1,Var2,Var3,Var4,Var5,Var6,Var7)]
Table2 <- in_table[, .(Pol_count = length(unique(POL))), by = list(Var2,Var3,Var4,Var5,Var6,Var7)]
Table3 <- in_table[, .(Pol_count = length(unique(POL))), by = list(Var1,Var3,Var4,Var5,Var6,Var7)]
and so on.
在表格中：
POL Var1 Var2 Var3 Var4 Var5 Var6 Var7
8035 1 11 1 GRD 0030 0110 09/30
36763 18813 GRD 5260 0300 11/15
36763 18813 GRD 5280 0300 11/15
3578618813 GRD 003001010 09/30
梳齿台：
Var1 Var2 Var3 Var4 Var5 Var6 Var7
表1
表2 0 1 1
表311
表1可能是这样的：
创建一个变量名为1
且NA
为0
nm_list <- data.frame( do.call("rbind", Map( function(x,y) as.character(factor(x, levels = c(0,1), labels = c(NA, y))),
x = Comb_table, y = names(Comb_table))),
stringsAsFactors = FALSE )
nm_list
# X1 X2 X3
# Var1 Var1 <NA> Var1
# Var2 Var2 Var2 <NA>
# Var3 Var3 Var3 Var3
# Var4 Var4 Var4 Var4
# Var5 Var5 Var5 Var5
# Var6 Var6 Var6 Var6
# Var7 Var7 Var7 Var7
library('data.table')
setDT(in_table) # convert data frame to data table by reference
lapply( nm_list, function(x) {
x <- na.omit(x) # remove NA
in_table[, .(Pol_count = length(unique(POL))), by = x] # extract the variables by passing the values to by argument
})
# $X1
# Var1 Var2 Var3 Var4 Var5 Var6 Var7 Pol_count
# 1: 1 11 1 GRD 30 110 09/30 1
# 2: 1 88 13 GRD 5260 300 11/15 1
# 3: 1 88 13 GRD 5280 300 11/15 1
# 4: 1 88 13 GRD 30 110 09/30 1
#
# $X2
# Var2 Var3 Var4 Var5 Var6 Var7 Pol_count
# 1: 11 1 GRD 30 110 09/30 1
# 2: 88 13 GRD 5260 300 11/15 1
# 3: 88 13 GRD 5280 300 11/15 1
# 4: 88 13 GRD 30 110 09/30 1
#
# $X3
# Var1 Var3 Var4 Var5 Var6 Var7 Pol_count
# 1: 1 1 GRD 30 110 09/30 1
# 2: 1 13 GRD 5260 300 11/15 1
# 3: 1 13 GRD 5280 300 11/15 1
# 4: 1 13 GRD 30 110 09/30 1
nm\u列表此功能：
> library(magrittr)
> melt(comb_table, id="tab", variable.factor=FALSE)[value == 1] %>%
split(by="tab") %>%
lapply(function(z) in_table[, .(n = uniqueN(POL)), by=c(z$variable)])
$Table1
Var1 Var2 Var3 Var4 Var5 Var6 Var7 n
1: 1 11 1 GRD 30 110 09/30 1
2: 1 88 13 GRD 5260 300 11/15 1
3: 1 88 13 GRD 5280 300 11/15 1
4: 1 88 13 GRD 30 110 09/30 1
$Table3
Var1 Var3 Var4 Var5 Var6 Var7 n
1: 1 1 GRD 30 110 09/30 1
2: 1 13 GRD 5260 300 11/15 1
3: 1 13 GRD 5280 300 11/15 1
4: 1 13 GRD 30 110 09/30 1
$Table2
Var2 Var3 Var4 Var5 Var6 Var7 n
1: 11 1 GRD 30 110 09/30 1
2: 88 13 GRD 5260 300 11/15 1
3: 88 13 GRD 5280 300 11/15 1
4: 88 13 GRD 30 110 09/30 1
这里使用magrittr只是为了方便
或者，如果您可以将所有数据都放在一个表中，并且正在使用data.table>=1.10.5，那么类似的内容（我没有测试过它…）应该可以用于分组集：
> melt(comb_table, id="tab", variable.factor=FALSE)[value == 1, groupingsets(
in_table,
sets = split(variable, tab)
)]
使用的数据：我决定OP的行名是/应该是一个名为“tab”的列
res=comb_table[，（列表（在_table[，uniqueN（POL），by=c（name（.SD）[.SD==1]）），by=tab]
#表V1
#1：表1
#2：表2
#3：表3
res$V1
#[[1]]
#Var1 Var2 Var3 Var4 Var5 Var6 Var7 V1
#1:1111 GRD311009/3011
#2:18813GRD52603001/1511
#3:18813GRD528030011/1511
#4:18813GRD311009/3011
#
#[[2]]
#Var2 Var3 Var4 Var5 Var6 Var7 V1
#1:11 GRD311009/3011
#2:88 13 GRD 5260 300 11/15 1
#3:88 13 GRD 5280 300 11/15 1
#4:8813GRD311009/3011
#
#[[3]]
#Var1 Var3 Var4 Var5 Var6 Var7 V1
#1:11 GRD 30110 09/30 1
#2:113GRD526030011/1511
#3:113GRD528030011/1511
#4:113GRD311009/3011
可能会看到？uniqueN
和？分组集
。下面是后者的一个例子
> melt(comb_table, id="tab", variable.factor=FALSE)[value == 1, groupingsets(
in_table,
sets = split(variable, tab)
)]
> dput(setDF(comb_table))
structure(list(tab = c("Table1", "Table2", "Table3"), Var1 = c(1L,
0L, 1L), Var2 = c(1L, 1L, 0L), Var3 = c(1L, 1L, 1L), Var4 = c(1L,
1L, 1L), Var5 = c(1L, 1L, 1L), Var6 = c(1L, 1L, 1L), Var7 = c(1L,
1L, 1L)), .Names = c("tab", "Var1", "Var2", "Var3", "Var4", "Var5",
"Var6", "Var7"), row.names = c(NA, -3L), class = "data.frame")
> dput(setDF(in_table))
structure(list(POL = c(8035L, 36763L, 36763L, 35786L), Var1 = c(1L,
1L, 1L, 1L), Var2 = c(11L, 88L, 88L, 88L), Var3 = c(1L, 13L,
13L, 13L), Var4 = c("GRD", "GRD", "GRD", "GRD"), Var5 = c(30L,
5260L, 5280L, 30L), Var6 = c(110L, 300L, 300L, 110L), Var7 = c("09/30",
"11/15", "11/15", "09/30")), .Names = c("POL", "Var1", "Var2",
"Var3", "Var4", "Var5", "Var6", "Var7"), row.names = c(NA, -4L
), class = "data.frame")
res = comb_table[, .(list(in_table[, uniqueN(POL), by = c(names(.SD)[.SD==1])])), by = tab]
# tab V1
#1: Table1 <data.table>
#2: Table2 <data.table>
#3: Table3 <data.table>
res$V1
#[[1]]
# Var1 Var2 Var3 Var4 Var5 Var6 Var7 V1
#1: 1 11 1 GRD 30 110 09/30 1
#2: 1 88 13 GRD 5260 300 11/15 1
#3: 1 88 13 GRD 5280 300 11/15 1
#4: 1 88 13 GRD 30 110 09/30 1
#
#[[2]]
# Var2 Var3 Var4 Var5 Var6 Var7 V1
#1: 11 1 GRD 30 110 09/30 1
#2: 88 13 GRD 5260 300 11/15 1
#3: 88 13 GRD 5280 300 11/15 1
#4: 88 13 GRD 30 110 09/30 1
#
#[[3]]
# Var1 Var3 Var4 Var5 Var6 Var7 V1
#1: 1 1 GRD 30 110 09/30 1
#2: 1 13 GRD 5260 300 11/15 1
#3: 1 13 GRD 5280 300 11/15 1
#4: 1 13 GRD 30 110 09/30 1