r - 仅选择数据框中与 R 中另一个数据框具有相同列名的列

Question

我有几个看起来像这样的数据集:

df1 <- data.frame (
A1_01 = c(1, 0, 0, 1, 0, 1, 0, 1, 0, 0),
A2_01 = c(1, 1, 1, 0, 1, 0, 0, 0, 0, 0),
A3_02 = c(0, 0, 0, 1, 0, 1, 0, 1, 1, 0),
L1_02 = c(1, 1, 1, 1, 1, 0, 0, 1, 1, 0),
L2_02 = c(0, 0, 0, 1, 1, 1, 0, 1, 0, 0),
age = rep(c("40-44", "45-49", "50-54", "55-59", "60-64"),2),
gender = c(rep("M",5), rep("F",5)),
ID = c("A12345", "A23456", "A34767", "A34567", "A45678", "A67891", "A78910", "A91011", 
     "A10111", "A11121"))

df2 <- data.frame (
A1_01 = c(1, 0, 0, 1, 0, 1, 0, 1, 0, 0),
A2_01 = c(1, 1, 1, 0, 1, 0, 0, 0, 0, 0),
A3_02 = c(0, 0, 0, 1, 0, 1, 0, 1, 1, 0),
Z4_02 = c(1, 1, 1, 1, 1, 0, 0, 1, 1, 0),
Z5_02 = c(0, 0, 0, 1, 1, 1, 0, 1, 0, 0),
age = rep(c("40-44", "45-49", "50-54", "55-59", "60-64"),2),
gender = c(rep("M",5), rep("F",5)),
ID = c("Q12345", "Q23456", "Q34767", "Q34567", "Q45678", "Q67891", "Q78910", "Q91011", 
     "Q10111", "Q11121"))

我想将所有这些数据集绑定(bind)在一起，形成一个更大的数据集。为此，我需要每个数据集都具有相同的列名。因此，我尝试对所有数据集进行子集化，使其只包含它们共有的列/变量。

这是我尝试做的，但这不起作用。

test <- df1 %>%
 select(names(df1) %in% names(df2))

我想要的输出是:

df3 <- data.frame (
A1_01 = c(1, 0, 0, 1, 0, 1, 0, 1, 0, 0, 1, 0, 0, 1, 0, 1, 0, 1, 0, 0),
A2_01 = c(1, 1, 1, 0, 1, 0, 0, 0, 0, 0, 1, 1, 1, 0, 1, 0, 0, 0, 0, 0),
A3_02 = c(0, 0, 0, 1, 0, 1, 0, 1, 1, 0, 0, 0, 0, 1, 0, 1, 0, 1, 1, 0),
age = c(rep(c("40-44", "45-49", "50-54", "55-59", "60-64"),2), rep(c("40-44", "45-49", "50-54", "55-59", "60-64"),2)),
gender = c(rep("M",5), rep("F",5), rep("M",5), rep("F",5)),
ID = c("A12345", "A23456", "A34767", "A34567", "A45678", "A67891", "A78910", "A91011", 
     "A10111", "A11121", "Q12345", "Q23456", "Q34767", "Q34567", "Q45678", "Q67891", "Q78910", "Q91011", 
     "Q10111", "Q11121") )

根据以下回复，我的代码现在很长。因为我有多个数据集，所以这非常耗时。本练习的目的是仅对所有六个数据集中共有的列进行子集化，因此我不想使用 bind_rows。

我最终编写了如下所示的迭代代码。有谁知道是否有更有效的方法来做到这一点？谢谢。

nm = intersect(names(NZHS_Y2A), names(NZHS_Y3A))
NZHS_Y2_3 <- rbind(NZHS_Y2A[nm], NZHS_Y3A[nm])
nm = intersect(names(NZHS_Y3A), names(NZHS_Y4A))
NZHS_Y3_4 <- rbind(NZHS_Y3A[nm], NZHS_Y4A[nm])
nm = intersect(names(NZHS_Y4A), names(NZHS_Y5A))
NZHS_Y4_5 <- rbind(NZHS_Y4A[nm], NZHS_Y5A[nm])
nm = intersect(names(NZHS_Y5A), names(NZHS_Y6A))
NZHS_Y5_6 <- rbind(NZHS_Y5A[nm], NZHS_Y6A[nm])

nm = intersect(names(NZHS_Y2_3), names(NZHS_Y3_4))
NZHS_Y2_4 <- rbind(NZHS_Y2_3[nm], NZHS_Y3_4[nm])
nm = intersect(names(NZHS_Y3_4), names(NZHS_Y4_5))
NZHS_Y3_5 <- rbind(NZHS_Y3_4[nm], NZHS_Y4_5[nm])
nm = intersect(names(NZHS_Y4_5), names(NZHS_Y5_6))
NZHS_Y4_6 <- rbind(NZHS_Y4_5[nm], NZHS_Y5_6[nm])

nm = intersect(names(NZHS_Y2_4), names(NZHS_Y3_5))
NZHS_Y2_5 <- rbind(NZHS_Y2_4[nm], NZHS_Y3_5[nm])
nm = intersect(names(NZHS_Y3_5), names(NZHS_Y4_6))
NZHS_Y3_6 <- rbind(NZHS_Y3_5[nm], NZHS_Y4_6[nm])

nm = intersect(names(NZHS_Y2_5), names(NZHS_Y4_6))
NZHS_Ad_2_6 <- rbind(NZHS_Y2_5[nm], NZHS_Y4_6[nm])

Answer 1

您可以使用 intersect 获取两个数据帧之间共有的列集，如 comment by d.b 中所述。

另一种方法是使用 dplyr 的 bind_rows，它允许您匹配匹配的列并用缺失值填充那些不匹配的列。在某些情况下，这可能是更理想的输出。

编辑:要处理许多数据帧，您应该将它们存储在列表中并使用 reduce 获取所有数据帧的交集。这会将函数应用于列表中的前两个元素，然后是该元素的结果和第三个元素，依此类推。然后，您可以在列表上 map_dfr 以仅选择每个数据帧中的共享列并将它们行绑定(bind)在一起(或者 map 然后 do.call(rbind , .) 如果你想使用 rbind。在这种情况下不需要，但是 bind_rows 直接接受一个列表作为输入。

df1 <- data.frame(
  A1_01 = c(1, 0, 0, 1, 0, 1, 0, 1, 0, 0),
  A2_01 = c(1, 1, 1, 0, 1, 0, 0, 0, 0, 0),
  A3_02 = c(0, 0, 0, 1, 0, 1, 0, 1, 1, 0),
  L1_02 = c(1, 1, 1, 1, 1, 0, 0, 1, 1, 0),
  L2_02 = c(0, 0, 0, 1, 1, 1, 0, 1, 0, 0),
  age = rep(c("40-44", "45-49", "50-54", "55-59", "60-64"), 2),
  gender = c(rep("M", 5), rep("F", 5)),
  ID = c(
    "A12345", "A23456", "A34767", "A34567", "A45678", "A67891", "A78910", "A91011",
    "A10111", "A11121"
  )
)

df2 <- data.frame(
  A1_01 = c(1, 0, 0, 1, 0, 1, 0, 1, 0, 0),
  A2_01 = c(1, 1, 1, 0, 1, 0, 0, 0, 0, 0),
  A3_02 = c(0, 0, 0, 1, 0, 1, 0, 1, 1, 0),
  Z4_02 = c(1, 1, 1, 1, 1, 0, 0, 1, 1, 0),
  Z5_02 = c(0, 0, 0, 1, 1, 1, 0, 1, 0, 0),
  age = rep(c("40-44", "45-49", "50-54", "55-59", "60-64"), 2),
  gender = c(rep("M", 5), rep("F", 5)),
  ID = c(
    "Q12345", "Q23456", "Q34767", "Q34567", "Q45678", "Q67891", "Q78910", "Q91011",
    "Q10111", "Q11121"
  )
)
library(tidyverse)
df_list <- list(df1, df2)
cols <- reduce(df_list, .f = ~ intersect(colnames(.x), colnames(.y)))
map_dfr(df_list, ~ .[cols])
#>    A1_01 A2_01 A3_02   age gender     ID
#> 1      1     1     0 40-44      M A12345
#> 2      0     1     0 45-49      M A23456
#> 3      0     1     0 50-54      M A34767
#> 4      1     0     1 55-59      M A34567
#> 5      0     1     0 60-64      M A45678
#> 6      1     0     1 40-44      F A67891
#> 7      0     0     0 45-49      F A78910
#> 8      1     0     1 50-54      F A91011
#> 9      0     0     1 55-59      F A10111
#> 10     0     0     0 60-64      F A11121
#> 11     1     1     0 40-44      M Q12345
#> 12     0     1     0 45-49      M Q23456
#> 13     0     1     0 50-54      M Q34767
#> 14     1     0     1 55-59      M Q34567
#> 15     0     1     0 60-64      M Q45678
#> 16     1     0     1 40-44      F Q67891
#> 17     0     0     0 45-49      F Q78910
#> 18     1     0     1 50-54      F Q91011
#> 19     0     0     1 55-59      F Q10111
#> 20     0     0     0 60-64      F Q11121
bind_rows(df_list)
#>    A1_01 A2_01 A3_02 L1_02 L2_02   age gender     ID Z4_02 Z5_02
#> 1      1     1     0     1     0 40-44      M A12345    NA    NA
#> 2      0     1     0     1     0 45-49      M A23456    NA    NA
#> 3      0     1     0     1     0 50-54      M A34767    NA    NA
#> 4      1     0     1     1     1 55-59      M A34567    NA    NA
#> 5      0     1     0     1     1 60-64      M A45678    NA    NA
#> 6      1     0     1     0     1 40-44      F A67891    NA    NA
#> 7      0     0     0     0     0 45-49      F A78910    NA    NA
#> 8      1     0     1     1     1 50-54      F A91011    NA    NA
#> 9      0     0     1     1     0 55-59      F A10111    NA    NA
#> 10     0     0     0     0     0 60-64      F A11121    NA    NA
#> 11     1     1     0    NA    NA 40-44      M Q12345     1     0
#> 12     0     1     0    NA    NA 45-49      M Q23456     1     0
#> 13     0     1     0    NA    NA 50-54      M Q34767     1     0
#> 14     1     0     1    NA    NA 55-59      M Q34567     1     1
#> 15     0     1     0    NA    NA 60-64      M Q45678     1     1
#> 16     1     0     1    NA    NA 40-44      F Q67891     0     1
#> 17     0     0     0    NA    NA 45-49      F Q78910     0     0
#> 18     1     0     1    NA    NA 50-54      F Q91011     1     1
#> 19     0     0     1    NA    NA 55-59      F Q10111     1     0
#> 20     0     0     0    NA    NA 60-64      F Q11121     0     0

由 reprex package (v0.2.0) 创建于 2018-08-01。