在 R 的循环中的各种数据帧中创建变量？

Question

我有 n 个数据框，我想为我的每个数据框创建一个相同的变量，该变量填充了数据框的相应名称。

i.e.:

dataframe_a

var 1 | var 2 | var 3 | variable
  x   |   x   |   x   | dataframe_a
  x   |   x   |   x   | dataframe_a
  x   |   x   |   x   | dataframe_a


dataframe_b

var 1 | var 2 | var 3 | variable
  x   |   x   |   x   | dataframe_b
  x   |   x   |   x   | dataframe_b
  x   |   x   |   x   | dataframe_b


dataframe_c

var 1 | var 2 | var 3 | variable
  x   |   x   |   x   | dataframe_c
  x   |   x   |   x   | dataframe_c
  x   |   x   |   x   | dataframe_c

有两种方法（如下），我可以分别为每个数据帧做我想做的事。但是，我怎样才能把下面的代码变成一个for循环，这样我就不用写100次了呢？

（我正在使用 dplyr::mutate）

#Take a, b and c as separate dataframes 

dataframe_a$variable <- paste("dataframe_a")
dataframe_b$variable <- paste("dataframe_b")
dataframe_c$variable <- paste("dataframe_c")

#or

dataframe_a %<>% mutate(variable = paste("dataframe_a"))
dataframe_b %<>% mutate(variable = paste("dataframe_b"))
dataframe_c %<>% mutate(variable = paste("dataframe_c"))

我尝试了以下方法，但它不起作用

vector <- c("a", "b", "c") 

for (i in vector){

  df <- get(i)

  assign(i, mutate(location = paste(i)), envir = .GlobalEnv)
}

Answer 1

最好在列表中使用它们。使用ls 和pattern 将所有数据框收集在一起，然后使用mget 获取实际数据框，然后使用cbind。

names_df <- ls(pattern = "dataframe")
Map(cbind, mget(names_df), variable = names_df)

使用可重现的数据

dataframe_a <- data.frame(a = 1:5, b = 6:10)
dataframe_b <- data.frame(a = 1:5, b = 6:10)
dataframe_c <- data.frame(a = 1:5, b = 6:10)

names_df <- ls(pattern = "dataframe")
df1 <- Map(cbind, mget(names_df), variable = names_df)
df1

#$dataframe_a
#  a  b    variable
#1 1  6 dataframe_a
#2 2  7 dataframe_a
#3 3  8 dataframe_a
#4 4  9 dataframe_a
#5 5 10 dataframe_a

#$dataframe_b
#  a  b    variable
#1 1  6 dataframe_b
#2 2  7 dataframe_b
#3 3  8 dataframe_b
#4 4  9 dataframe_b
#5 5 10 dataframe_b

#$dataframe_c
#  a  b    variable
#1 1  6 dataframe_c
#2 2  7 dataframe_c
#3 3  8 dataframe_c
#4 4  9 dataframe_c
#5 5 10 dataframe_c

现在您可以通过df1[["dataframe_a"]]、df1[["dataframe_b"]] 等访问单个数据帧。

Answer 2

我们可以使用imap

library(tidyverse)
imap(mget(names_df), ~ .x %>%
            mutate(variable = .y))
#$dataframe_a
#  a  b    variable
#1 1  6 dataframe_a
#2 2  7 dataframe_a
#3 3  8 dataframe_a
#4 4  9 dataframe_a
#5 5 10 dataframe_a

#$dataframe_b
#  a  b    variable
#1 1  6 dataframe_b
#2 2  7 dataframe_b
#3 3  8 dataframe_b
#4 4  9 dataframe_b
#5 5 10 dataframe_b

#$dataframe_c
#  a  b    variable
#1 1  6 dataframe_c
#2 2  7 dataframe_c
#3 3  8 dataframe_c
#4 4  9 dataframe_c
#5 5 10 dataframe_c

数据

dataframe_a <- data.frame(a = 1:5, b = 6:10)
dataframe_b <- data.frame(a = 1:5, b = 6:10)
dataframe_c <- data.frame(a = 1:5, b = 6:10)

names_df <- ls(pattern = "dataframe")