Replace values in data.frame for multiple columns and merge

Question

Given the following data.frame df:

And want to replace the values in each column where the value is set to 1 (is there a smarter possibility as the following?):

df["Typ1"][df["Typ1"] == 1]<-"Typ1"
df["Typ2"][df["Typ2"] == 1]<-"Typ2"

And merge the columns to:

   Typ 
  "Typ1"           
  "Typ2"           
  "Typ1"

score 3 · Accepted Answer · answered Oct 21 '22 at 18:18

library(tidyr)
df %>% 
  pivot_longer(Typ1:Typ2, names_to = "Typ") %>% 
  filter(value == 1) %>% 
  select(Typ)

Output:

# A tibble: 3 × 1
  Typ  
  <chr>
1 Typ1 
2 Typ2 
3 Typ1

Input:

df <- structure(list(Typ1 = c("1", "0", "1"), Typ2 = c("0", "1", "0"
)), row.names = 2:4, class = "data.frame")

score 0 · Answer 2 · answered Oct 21 '22 at 18:28

0

Data:

Typ1 <- sample(rep(c(1,0),10))
Typ2 <- as.numeric(Typ1<1)
df <- data.frame(Typ1,Typ2)

Slightly more efficient indexing solution:

df$Typ <- NA
df$Typ[df$Typ2 == 1] <- "Typ2"
df$Typ[df$Typ1 == 1] <- "Typ1"

answered Oct 21 '22 at 18:28

Roasty247

This is what the OP wants to avoid – Jilber Urbina Oct 21 '22 at 18:34
I don't think they said they want to avoid indexing. Just "Is there a smarter possibilty", this only uses 1 set of `[ ]` rather than 3 sets of `[ ]` in OP ...so an improvement no? – Roasty247 Oct 21 '22 at 18:36

2 Answers2