r - 从非因子列生成键列

Question

给定一个数据框的形式

       Key.1 Key.2      Value
1  5/25/2018   -10 0.53928999
2  5/25/2018   -10 0.23083204
3  5/25/2018   -10 0.33742676
4  5/25/2018     0 0.53479860
5  5/25/2018     0 0.27612761
6  5/25/2018     0 0.74993199
7  5/25/2018    10 0.01397069
8  5/25/2018    10 0.10553610
9  5/25/2018    10 0.66147883
10 1/17/2018   -10 0.14381738
11 1/17/2018   -10 0.52708544
12 1/17/2018   -10 0.75862925
13 1/17/2018     0 0.45954116
14 1/17/2018     0 0.68467543
15 1/17/2018     0 0.15865298
16 1/17/2018    10 0.01039363
17 1/17/2018    10 0.49886623
18 1/17/2018    10 0.98269967
19 5/25/2018    10 0.10553610
20 5/25/2018   -10 0.33742676

我需要生成一个 Group来自 key.1 互动专栏和 key.2那看起来像

       Key.1 Key.2      Value Group
1  5/25/2018   -10 0.53928999     1
2  5/25/2018   -10 0.23083204     1
3  5/25/2018   -10 0.33742676     1
4  5/25/2018     0 0.53479860     2
5  5/25/2018     0 0.27612761     2
6  5/25/2018     0 0.74993199     2
7  5/25/2018    10 0.01397069     3
8  5/25/2018    10 0.10553610     3
9  5/25/2018    10 0.66147883     3
10 1/17/2018   -10 0.14381738     4
11 1/17/2018   -10 0.52708544     4
12 1/17/2018   -10 0.75862925     4
13 1/17/2018     0 0.45954116     5
14 1/17/2018     0 0.68467543     5
15 1/17/2018     0 0.15865298     5
16 1/17/2018    10 0.01039363     6
17 1/17/2018    10 0.49886623     6
18 1/17/2018    10 0.98269967     6
19 5/25/2018    10 0.10553610     3
20 5/25/2018   -10 0.33742676     1

注意最后两行

重要的是 Group 的值正在沿数据框上升。我已经设法获得所需的行为

Data$Group          <-  interaction(paste(Data$Key.1,Data$Key.2),1)
levels(Data$Group)  <-  1:length(levels(Data$Group))
levels(Data$Group)  <-  unique(Data$Group)

然而，这感觉非常不直观和笨拙。

如何使这既简短又直观？

需要注意的是，对 Key.1 的内容没有实际限制。或 Key.2可能是 - 核心行为只需是 Group由唯一的一对 Key.1 定义和 Key.2 ，然后从 table 上下来。

Answer 1

这是一个使用因子的想法:

使用基础 R:

df$Group = as.integer(factor(paste(df$Key.1, df$Key.2), 
                             levels = unique(paste(df$Key.1, df$Key.2))))

或与 mutate来自 dplyr :

library(dplyr)

df = mutate(df, Group = paste(Key.1, Key.2) %>% 
         factor(., levels = unique(.)) %>%
         as.integer())

结果:

       Key.1 Key.2      Value Group
1  5/25/2018   -10 0.53928999     1
2  5/25/2018   -10 0.23083204     1
3  5/25/2018   -10 0.33742676     1
4  5/25/2018     0 0.53479860     2
5  5/25/2018     0 0.27612761     2
6  5/25/2018     0 0.74993199     2
7  5/25/2018    10 0.01397069     3
8  5/25/2018    10 0.10553610     3
9  5/25/2018    10 0.66147883     3
10 1/17/2018   -10 0.14381738     4
11 1/17/2018   -10 0.52708544     4
12 1/17/2018   -10 0.75862925     4
13 1/17/2018     0 0.45954116     5
14 1/17/2018     0 0.68467543     5
15 1/17/2018     0 0.15865298     5
16 1/17/2018    10 0.01039363     6
17 1/17/2018    10 0.49886623     6
18 1/17/2018    10 0.98269967     6
19 5/25/2018    10 0.10553610     3
20 5/25/2018   -10 0.33742676     1

数据:

df = structure(list(Key.1 = c("5/25/2018", "5/25/2018", "5/25/2018", 
"5/25/2018", "5/25/2018", "5/25/2018", "5/25/2018", "5/25/2018", 
"5/25/2018", "1/17/2018", "1/17/2018", "1/17/2018", "1/17/2018", 
"1/17/2018", "1/17/2018", "1/17/2018", "1/17/2018", "1/17/2018", 
"5/25/2018", "5/25/2018"), Key.2 = c(-10L, -10L, -10L, 0L, 0L, 
0L, 10L, 10L, 10L, -10L, -10L, -10L, 0L, 0L, 0L, 10L, 10L, 10L, 
10L, -10L), Value = c(0.53928999, 0.23083204, 0.33742676, 0.5347986, 
0.27612761, 0.74993199, 0.01397069, 0.1055361, 0.66147883, 0.14381738, 
0.52708544, 0.75862925, 0.45954116, 0.68467543, 0.15865298, 0.01039363, 
0.49886623, 0.98269967, 0.1055361, 0.33742676)), .Names = c("Key.1", 
"Key.2", "Value"), class = "data.frame", row.names = c("1", "2", 
"3", "4", "5", "6", "7", "8", "9", "10", "11", "12", "13", "14", 
"15", "16", "17", "18", "19", "20"))