CODEWITHSUNDEEP
rsummaryweighted
Zerig
Zerig

Reputation: 43

R: weighted summary

I have survey, where people sort animals by their popularity. for example:

"first","second","third"
"dog","cat","rabbit"
"cat","rabbit","dog"
"dog","cat","rabbit"
"dog","cat","rabbit"

I want graph, which shows order. So animals in first column get weight 3, in second weight 2 and in third weight 1.

When i do:

data<-read.csv("animals.csv", header=T)

f = summary(data$first)
s = summary(data$second)
t = summary(data$third)

In f is array with:

dog      cat 
  3        1

In s is array with:

cat      rabbit
  3           1

In t is array with:

dog      rabbit
  1           3

And i obviously cannot do:

graph = 3*f + 2*s + t

Because they dont have the same column. Exist some function how to do it. How to make weighted combination of these three columns?

In the end i need:

dog      cat     rabbit
 10        9          5

I am a begginer in R, so thank for help.

Upvotes: 0

Views: 930

Answers (1)

Cath
Cath

Reputation: 24074

To get the weighted numbers, you can do:

table(c(rep(data$first, 3), rep(data$second, 2), data$third))

#   cat    dog rabbit 
#    9     10      5

data

data <- structure(list(first = c("dog", "cat", "dog", "dog"), second = c("cat", 
"rabbit", "cat", "cat"), third = c("rabbit", "dog", "rabbit", 
"rabbit")), .Names = c("first", "second", "third"), class = "data.frame", row.names = c(NA, 
-4L))

NB: if the data are imported with characters as factors, you can do:

table(c(rep(as.character(data$first), 3), rep(as.character(data$second), 2), as.character(data$third)))

Upvotes: 3

Related Questions