R/dplyr: combining two columns into and keeping each suborder

I have a dataframe

df <- data.frame(
  cities = c(
    rep("Atlanta", 3),
    rep("Seattle", 4),
    rep("Paris", 2)
  ),
  area = c(
    c("A", "C", "X"),
    c("B", "H", "I", "D"),
    c("Z", "F")
  )
)

and I would to create a new dataframe with one column that represents a joined column of the columns cities and area. This is the output I would like to obtain:

data.frame(
  new_column = c(
    "Atlanta",
    "A",
    "C",
    "X",
    "Paris",
    "F",
    "Z",
    "Seattle",
    "B",
    "D",
    "H",
    "I"
  )
)

i.e. I want to keep the column cities and area alphabetically within each category but order that into one column in a new dataframe.

Is there an elegant way with dplyr to do that?

Thank you!

>Solution :

This might be good enough for you:

library(dplyr)
group_by(df, cities) %>%
  reframe(new_column = c(first(cities), sort(area)))
# # A tibble: 12 × 2
#    cities  new_column
#    <chr>   <chr>     
#  1 Atlanta Atlanta   
#  2 Atlanta A         
#  3 Atlanta C         
#  4 Atlanta X         
#  5 Paris   Paris     
#  6 Paris   F         
#  7 Paris   Z         
#  8 Seattle Seattle   
#  9 Seattle B         
# 10 Seattle D         
# 11 Seattle H         
# 12 Seattle I         

(and then remove the cities column).