CODEWITHSUNDEEP
rpercentage
user3077008
user3077008

Reputation: 847

Calculating a row percentage for a specific value in R

I've been trying to find a way to calculate the proportion of 1 count in each row in R.
For example, in following data. 15090 is have one 1 in 5 votes, so his proportion is 0.2 In the next row, he has four 1s out of 5 votes, so the proportion is 0.8. I need to make NA to be dismissed in the calculation, i.e. should not be included in the denominator. 

ICPSR         Vote.335 Vote.354 Vote.971 Vote.972 Vote.973 Vote.155 Vote.295 
15090         0        0        0        1        0        NA       NA
29300         1        1        1        0        1        NA       NA

What code should in R, to calculate each row's proportion of 1?
Thank you in advance!

Upvotes: 0

Views: 2279

Answers (3)

Sven Hohenstein
Sven Hohenstein

Reputation: 81743

If the columns contain 0s and 1s only, you can use

rowMeans(dat[-1], na.rm = TRUE)
#[1] 0.2 0.8

where dat is the name of your data frame.

Upvotes: 1

agstudy
agstudy

Reputation: 121608

Here a vectorized solution using rowSums:

dat <- dat[,-1]
rowSums(dat==1,na.rm=TRUE)/rowSums(!is.na(dat))
 [1] 0.2 0.8

where dat:

dput(dat)

structure(list(ICPSR = c(15090L, 29300L), Vote.335 = 0:1, Vote.354 = 0:1, 
    Vote.971 = 0:1, Vote.972 = c(1L, 0L), Vote.973 = 0:1, Vote.155 = c(NA, 
    NA), Vote.295 = c(NA, NA)), .Names = c("ICPSR", "Vote.335", 
"Vote.354", "Vote.971", "Vote.972", "Vote.973", "Vote.155", "Vote.295"
), class = "data.frame", row.names = c(NA, -2L))

Upvotes: 1

Neal Fultz
Neal Fultz

Reputation: 9696

prop.table() and addmargins() do that.

Upvotes: 2

Related Questions