Combine variables AND create a data frame with a column with the name of those original variables wi-CodePudding

I have a group of variables I created in my R environment:

x <- c("My name is Andrea and I live in Vancouver",
       "I work at a university")
y <- c("My name is Andrew and I live in New York",
       "I work at a hospital")
z <- c("My name is Alessia and I live in Rome",
       "I work for the government")

I want to convert these "character" variables into tibbles and assign as a variable to each tibble the name of the dataset (so in this case, the names would be x, y and z).

Example of the tibble I'd like:

# A tibble: 6 × 2
  value                                          name                           
  <chr>                                          <chr> 
1 My name is Andrea and I live in Vancouver       "x"
2 I work at a university                          "x"
3 My name is Andrew and I live in New York        "y"
4 I work at a hospital                            "y"
5 My name is Alessia and I live in Rome           "z"
6 I work for the government                       "z"

Now the code here: test <- as_tibble(c(x,y,z)) %>% mutate(name = c("x", "y", "z")) doesn't work because of a mismatch in sizes.

This code does, but isn't in "dplyr" format:

xyz.list <- Hmisc::llist(x, y ,z)
df <- do.call(rbind, unname(Map(cbind, source = names(xyz.list), xyz.list)))
df <- df %>% as_tibble()

My question: is there a way to "merge" or combine different datasets AND create a column with the name of the original merged dataset using dplyr?

CodePudding user response：

In base R:

stack(list(x = x, y = y, z= z))
                                     values ind
1 My name is Andrea and I live in Vancouver   x
2                    I work at a university   x
3  My name is Andrew and I live in New York   y
4                      I work at a hospital   y
5     My name is Alessia and I live in Rome   z
6                 I work for the government   z

in tidyverse:

library(tidyverse)
unnest(enframe(lst(x, y, z)), value)

# A tibble: 6 x 2
  name  value                                    
  <chr> <chr>                                    
1 x     My name is Andrea and I live in Vancouver
2 x     I work at a university                   
3 y     My name is Andrew and I live in New York 
4 y     I work at a hospital                     
5 z     My name is Alessia and I live in Rome    
6 z     I work for the government

CodePudding user response：

Another tidyverse option:

library(tidyverse)

lst(x, y, z) %>% 
  bind_rows() %>% 
  pivot_longer(everything()) %>% 
  arrange(name)

Or we could use map_df from purrr:

map_df(lst(x, y, z), ~ data.frame(value = .x), .id = "name")

Output

  name  value                                    
  <chr> <chr>                                    
1 x     My name is Andrea and I live in Vancouver
2 x     I work at a university                   
3 y     My name is Andrew and I live in New York 
4 y     I work at a hospital                     
5 z     My name is Alessia and I live in Rome    
6 z     I work for the government