CODEWITHSUNDEEP
rdata-science
Gustavo Rubia
Gustavo Rubia

Reputation: 3

How can I create new columns with the values of multiple old columns?

first time posting so if something is wrong please let me know, I have a dataframe in R that is divided in the following way:

location type amount produt
a x 10 p1
a x 20 p2
b x 50 p5
b y 100 p10

In the end I need to group the locations into a single line and make the "type" column to become new columns with the value of the "product" and "amount" column, just like this :

location A_P_X A_P_Y
a p1_10,p2_20
b p5_50 p10_100

I tried to make the new coluns using one hot encoding but I run into problem when tring to fill the new columns based on their original "type" value

Upvotes: 0

Views: 37

Answers (1)

Jon Spring
Jon Spring

Reputation: 66415

library(tidyverse)
df %>%
  unite("val", produt:amount, sep = "_") %>%
  mutate(type = paste0("A_P_", toupper(type))) %>%
  pivot_wider(names_from = type, values_from = val, 
              values_fn = list(val = ~paste(., collapse = ", ")))

result

# A tibble: 2 x 3
  location A_P_X        A_P_Y  
  <chr>    <chr>        <chr>  
1 a        p1_10, p2_20 NA     
2 b        p5_50        p10_100

Upvotes: 1

Related Questions