CODEWITHSUNDEEP
rdataframe
EGM8686
EGM8686

Reputation: 1572

Getting row-wise percentages for a dataframe

Given a dataframe like this:

id v1  v2  v3  v4
1  10  20  60  10
2  10  10  10  70
3  50  25  10  15

I would like to get their rowwise % like this:

id v1  v2  v3  v4  p1   p2   p3   p4
1  10  20  60  10  0.1  0.2  0.6  0.1
2  10  10  10  70  0.1  0.1  0.1  0.7
3  50  25  10  15  0.5  0.25 0.10 0.15

So I need to retain the original variables and create new ones that reflect their relative %, in reality the variables go all the way to 55 so I'm looking for a simple solution that does not require to calculate each value individually.

Upvotes: 0

Views: 40

Answers (2)

Maurits Evers
Maurits Evers

Reputation: 50668

Here is a tidyverse option

library(tidyverse)
left_join(df, df %>%
    gather(k, v, -id) %>%
    group_by(id) %>%
    mutate(frac = v / sum(v), v = NULL, k = str_replace(k, "v", "p")) %>%
    spread(k, frac))
#  id v1 v2 v3 v4  p1   p2  p3   p4
#1  1 10 20 60 10 0.1 0.20 0.6 0.10
#2  2 10 10 10 70 0.1 0.10 0.1 0.70
#3  3 50 25 10 15 0.5 0.25 0.1 0.15

Sample data

df <- read.table(text =
    "id v1  v2  v3  v4
1  10  20  60  10
2  10  10  10  70
3  50  25  10  15", header = T)

Upvotes: 0

d.b
d.b

Reputation: 32548

setNames(object = df1[,-1]/rowSums(df1[,-1]), 
         nm = gsub("[(v)\\d+]", "p", colnames(df1[,-1])))
#   p1   p2  p3   p4
#1 0.1 0.20 0.6 0.10
#2 0.1 0.10 0.1 0.70
#3 0.5 0.25 0.1 0.15

Upvotes: 2

Related Questions