CODEWITHSUNDEEP
r
Aryh
Aryh

Reputation: 595

How to get Percentage in two categorical variables in R

Just suppose you have total number of samples as 8. Data frame looks like that. All the individuals having Healthscore less than 3 are Healthy and All having health scores greater than 3 are Sick. Status shows their employement status.

Status<-(Employed,Unemployed,Student,Student,Employed,Unemployed,Unemployed,Housewife)
Health<-(Healthy,Healthy,Healthy,Sick,Sick,Control,Sick,Sick)

df<-(Status,Health)
level(Health)<-("Healthy,"Sick",Control)
level(Status)<-("Employed","Unemployed","Student","Housewife")

I want to see the percentage of Healthy,Sick or Control people belong to each occupation category. I want output like following. (p.s values are just hypothetical in example) like OUT OF ALL EMPLOYED INDIVIDUALS, HOW MANY PERCENTS ARE HEALTHY??

                    Healthy      Sick   Control
Employed              10%        2%     1%
Unemployed             5%        1%     1%
Student                6%        3%     1%
Housewife              2%        5%     6%

I am using following code. But it just gives me frequencies, NOT PERCENTAGE. I need percentage. 

tab <- with(df, table(df$Health,df$Status))

Upvotes: 0

Views: 1585

Answers (1)

Ronak Shah
Ronak Shah

Reputation: 388817

We can count the number of individuals for each Status and Health, group_by Status and calculate the percentage. For better visibility we cast the data in wide format. 

library(dplyr)

df %>%
  count(Status, Health) %>%
  group_by(Status) %>%
  mutate(n = n/sum(n) * 100) %>%
  tidyr::pivot_wider(names_from = Health, values_from = n, 
                     values_fill = list(n = 0))


# Status     Healthy  Sick Control
#  <fct>        <dbl> <dbl>   <dbl>
#1 Employed      50    50       0  
#2 Housewife      0   100       0  
#3 Student       50    50       0  
#4 Unemployed    33.3  33.3    33.3

In base R, we can use prop.table along with table to get the percentages. 

prop.table(table(df), 1) * 100

data

df <- structure(list(Status = structure(c(1L, 4L, 3L, 3L, 1L, 4L, 4L, 
2L), .Label = c("Employed", "Housewife", "Student", "Unemployed"
), class = "factor"), Health = structure(c(2L, 2L, 2L, 3L, 3L, 
1L, 3L, 3L), .Label = c("Control", "Healthy", "Sick"), 
class = "factor")), class = "data.frame",row.names = c(NA, -8L))

Upvotes: 1

Related Questions