将 R 中的单个 dataframe 中的多列合并为一个新的 dataframe 作为长格式的一列

Question

My dataframe looks like

ID  FMR  FML  KPR  KPL
----------------------
a    1   NA   NA   1
b    0   NA    1   0
c    NA   1   NA   NA
d    1   NA    0   NA
e    NA   1   NA   NA

I need to make a new dataframe that looks like

ID   FM   KP
-------------
a    1    NA
b    0     1
c    NA   NA
d    1     0
e    NA   NA
a    NA    1
b    NA    0
c    1    NA
d    NA   NA
e    1    NA

What would be the best way to do this?

Answer 1

tidyverse ( coalesce() specifically) is your friend here...

fm <- df %>%
  transmute(ID,
            FM = coalesce(FMR, FML))

kp <- df %>%
  transmute(ID,
            KP = coalesce(KPR, KPL))
fm %>%
  bind_rows(kp)

# A tibble: 10 x 3
   ID       FM    KP
   <chr> <dbl> <dbl>
 1 a         1    NA
 2 b         0    NA
 3 c         1    NA
 4 d         1    NA
 5 e         1    NA
 6 a        NA     1
 7 b        NA     1
 8 c        NA    NA
 9 d        NA     0
10 e        NA    NA

but are you sure you want that output? You would be losing information on whether KP is coming from KPR or KPL for example.

Answer 2

From the example shared it can be seen that the first two characters of column name identify which column a particular value goes. We can use that in tidyr 's pivot_longer .

tidyr::pivot_longer(df, cols = -ID, 
                    names_to = '.value', 
                    names_pattern = '(..)')


#   ID       FM    KP
#   <chr> <int> <int>
# 1 a         1    NA
# 2 a        NA     1
# 3 b         0     1
# 4 b        NA     0
# 5 c        NA    NA
# 6 c         1    NA
# 7 d         1     0
# 8 d        NA    NA
# 9 e        NA    NA
#10 e         1    NA

This will work for any number of columns as far as the column name pattern holds.

data

df <- structure(list(ID = c("a", "b", "c", "d", "e"), FMR = c(1L, 0L, 
NA, 1L, NA), FML = c(NA, NA, 1L, NA, 1L), KPR = c(NA, 1L, NA, 
0L, NA), KPL = c(1L, 0L, NA, NA, NA)), class = "data.frame", 
row.names = c(NA, -5L))