使用列表在 R 中使用 dplyr 创建新变量

Question

假设我们有三个列表。

list_A <- c("PA","MA","MD")
list_B <- c("NJ","NY","OK")
list_C <- c("AZ","MT","LA")

我有一个这样的数据框

ID        presenter          state   
1         Donatello   c("AZ","NY")
2          Leonardo             NJ
3            Rafael   c("LA","MT")
4     Michaelangelo    c("PA,"LA")

我想使用 tidyverse 创建三个新变量，计算 state 的每行包含其列表元素的次数。

ID        presenter          state     A   B   C   
1         Donatello   c("AZ","NY")     0   1   1
2          Leonardo             NJ     0   1   0
3            Rafael   c("LA","MT")     0   0   2
4     Michaelangelo   c("PA","LA")     1   0   1

不相关但相关，但出于好奇，是否可以 unlist() 状态但使用旧信息创建更多行，因此操纵 df1...

ID        presenter          state   
1         Donatello             AZ
1         Donatello             NY
2          Leonardo             NJ
3            Rafael             LA
3            Rafael             MT
4     Michaelangelo             PA
4     Michaelangelo             LA

Answer 1

你可以使用双sapply：

list_data <- list(list_A, list_B, list_C)
cbind(df1, data.frame(sapply(list_data, function(x) 
           sapply(df1$state, function(y) sum(y %in% x)))))


#     ID presenter     state        X1    X2    X3
#  <int> <chr>         <list>    <int> <int> <int>
#1     1 Donatello     <chr [2]>     0     1     1
#2     2 Leonardo      <chr [1]>     0     1     0
#3     3 Rafael        <chr [2]>     0     0     2
#4     4 Michaelangelo <chr [2]>     1     0     1

Answer 2

关于您的第一个问题如何：

library(dplyr)
library(tidyr)
library(tibble)

list_A <- c("PA","MA","MD")
list_B <- c("NJ","NY","OK")
list_C <- c("AZ","MT","LA")


data <- tibble(
  ID = c(1, 2, 3, 4),
  presenter = c("Donatello", "Leonardo", "Rafael", "Michaelangelo"),
  state = list(c("AZ", "NJ"), c("NJ"), c("LA", "MT"), c("PA", "LA"))
)

data <- data %>%
  rowwise() %>%
  mutate(A = sum(list_A %in% state),
         B = sum(list_B %in% state),
         C = sum(list_C %in% state))

有了这个输出：

> data
Source: local data frame [4 x 6]
Groups: <by row>

# A tibble: 4 x 6
     ID presenter     state         A     B     C
  <dbl> <chr>         <list>    <int> <int> <int>
1     1 Donatello     <chr [2]>     0     1     1
2     2 Leonardo      <chr [1]>     0     1     0
3     3 Rafael        <chr [2]>     0     0     2
4     4 Michaelangelo <chr [2]>     1     0     1

更新 至于你的第二个问题：tidyrunnest 函数会做到这一点。

> data %>%
+   unnest(state)
# A tibble: 7 x 3
     ID presenter     state
  <dbl> <chr>         <chr>
1     1 Donatello     AZ   
2     1 Donatello     NJ   
3     2 Leonardo      NJ   
4     3 Rafael        LA   
5     3 Rafael        MT   
6     4 Michaelangelo PA   
7     4 Michaelangelo LA 

UDPATE2

要计算 state 列中的多次出现，您需要一个额外的循环。这可能会做到（但建议您对其进行测试）：

data <- data %>%
  rowwise() %>%
  mutate(A = sum(unlist(lapply(list_A, function(x) sum(x == state)))),
         B = sum(unlist(lapply(list_B, function(x) sum(x == state)))),
         C = sum(unlist(lapply(list_C, function(x) sum(x == state)))))

对于这个数据：

data <- tibble(
  ID = c(1, 2, 3, 4),
  presenter = c("Donatello", "Leonardo", "Rafael", "Michaelangelo"),
  state = list(c("AZ", "NJ"), c("NJ"), c("LA", "MT", "MT", "MT"), c("PA", "PA", "LA")) 
)

我们预计 C 列和 A 列的第三行 (3x "MT") 和第四行 (2x "PA") 会有额外的计数：

> data
Source: local data frame [4 x 6]
Groups: <by row>

# A tibble: 4 x 6
     ID presenter     state         A     B     C
  <dbl> <chr>         <list>    <int> <int> <int>
1     1 Donatello     <chr [2]>     0     1     1
2     2 Leonardo      <chr [1]>     0     1     0
3     3 Rafael        <chr [4]>     0     0     4
4     4 Michaelangelo <chr [3]>     2     0     1