How to calculate category frequencies for all variables for each group separately with absolute and percentage values in R

Question

my dataset has such structure

structure(list(GR = c(1L, 1L, 1L, 1L, 2L, 2L, 2L, 2L), Skin..visible.mucous.membranes..hair..nails.1 = c("Normal", 
"Normal", "Normal", "Normal", "Normal", "Normal", "Normal", "Normal"
), The.lymph.nodes.1 = c("Normal", "Normal", "Normal", "Normal", 
"Normal", "Normal", "Normal", "Normal"), ENT.organs..respiratory.system.1 = c("Normal", 
"Normal", "Normal", "Normal", "Normal", "Normal", "Normal", "Normal"
), The.cardiovascular.system.1 = c("Normal", "Normal", "Normal", 
"Normal", "Normal", "Normal", "Normal", "Normal"), Gastrointestinal..including.assessment.of.spleen.and.liver.size..1 = c("Normal", 
"Normal", "Normal", "Normal", "Normal", "Normal", "Normal", "Normal"
), Genitourinary.sphere.1 = c("Normal", "Normal", "Normal", "Normal", 
"Normal", "Normal", "Normal", "Normal"), Nervous.system.1 = c("Normal", 
"Normal", "Normal", "Normal", "Normal", "Normal", "Normal", "Normal"
), Psychic.Sphere.1 = c("Normal", "Normal", "Normal", "Normal", 
"Normal", "Normal", "Normal", "Normal"), Skin..visible.mucous.membranes..hair..nails.2 = c("Normal", 
"Normal", "Normal", "Normal", "Normal", "Normal", "Normal", "Normal"
), The.lymph.nodes.2 = c("Normal", "Normal", "Normal", "Normal", 
"Normal", "Normal", "Normal", "Normal"), ENT.organs..respiratory.system.2 = c("Normal", 
"Normal", "Normal", "Normal", "Normal", "Normal", "Normal", "Normal"
), The.cardiovascular.system.2 = c("Normal", "Normal", "Normal", 
"Normal", "Normal", "Normal", "Normal", "Normal"), Gastrointestinal..including.assessment.of.spleen.and.liver.size..2 = c("Normal", 
"Normal", "Normal", "Normal", "Normal", "Normal", "Normal", "Normal"
), Genitourinary.sphere.2 = c("Normal", "Normal", "Normal", "Normal", 
"Normal", "Normal", "Normal", "Normal"), Nervous.system.2 = c("Normal", 
"Normal", "Normal", "Normal", "Normal", "Normal", "Normal", "Normal"
), Psychic.Sphere.2 = c("Normal", "Normal", "Normal", "Normal", 
"Normal", "Normal", "Normal", "Normal"), Skin..visible.mucous.membranes..hair..nails.3 = c("Normal", 
"Normal", "Normal", "Normal", "Normal", "Normal", "Normal", "Normal"
), The.lymph.nodes.3 = c("Normal", "Normal", "Normal", "Normal", 
"Normal", "Normal", "Normal", "Normal"), ENT.organs..respiratory.system.3 = c("Normal", 
"Normal", "Normal", "Normal", "Normal", "Normal", "Normal", "Normal"
), The.cardiovascular.system.3 = c("Normal", "Normal", "Normal", 
"Normal", "Normal", "Normal", "Normal", "Normal"), Gastrointestinal..including.assessment.of.spleen.and.liver.size..3 = c("Normal", 
"Normal", "Normal", "Normal", "Normal", "Normal", "Normal", "Normal"
), Genitourinary.sphere.3 = c("Normal", "Normal", "Normal", "Normal", 
"Normal", "Normal", "Normal", "Normal"), Nervous.system.3 = c("Normal", 
"Normal", "Normal", "Normal", "Normal", "Normal", "Normal", "Normal"
), Psychic.Sphere.3 = c("Normal", "Normal", "Normal", "Normal", 
"Normal", "Normal", "Normal", "Normal")), class = "data.frame", row.names = c(NA, 
-8L))

this is part of the results of the study on which i need to calculate the frequencies for the normal or not normal categories.

How to calculate frequencies for categories of normal and not normal for each of studied parameter for each group separately

Skin, visible mucous membranes, hair, nails
The lymph nodes
  ENT organs, respiratory system
The cardiovascular system
Gastrointestinal (including assessment of spleen and liver size)
Genitourinary sphere
Nervous system
Psychic Sphere

and each of them has a digital prefix, for example The lymph nodes 1 - means the first day of the study, The lymph nodes 2 - means the second day of the study. To make a reproducible example, I took only 3 days. And each parameter has, of course, GR study groups.

so that the desired final table looks like with absolute and percentage values

I know how to use the table command, but it does not allow me get desired output result in exactly the format that I showed in the figure.(this output i self create manually using full data and here it as example desired format)

How can I calculate frequencies like this to get needed desired output?

user_17973287 · Answer

You can try a code as follows:

library(tidyverse)

> df
# A tibble: 18 x 4
   Var      GR result      days
           
 1 skin      1 normal         1
 2 skin      1 normal         1
 3 skin      2 normal         1
 4 skin      2 normal         1
 5 skin      1 not_normal     1
 6 skin      2 not_normal     1
 7 skin      1 not_normal     1
 8 skin      1 normal         2
 9 skin      1 normal         2
10 skin      2 normal         2
11 skin      1 not_normal     2
12 skin      2 not_normal     2
13 skin      1 not_normal     3
14 skin      1 not_normal     3
15 skin      2 normal         3
16 skin      2 normal         3
17 skin      2 normal         3
18 skin      2 not_normal     3


df %>% group_by(Var, GR,days, result) %>% summarise(n = n()) %>% mutate(p = round(100 * n / sum(n, na.rm = T),2), rep = paste0( n, " (", p,"%)")) %>% select(-n , - p)  %>% pivot_wider(names_from = days, names_prefix = "day_", values_from = rep) %>% arrange( Var, GR, result)

# A tibble: 4 x 6
# Groups:   Var, GR [2]
  Var      GR result     day_1      day_2      day_3   
                         
1 skin      1 normal     2 (50%)    2 (66.67%) NA      
2 skin      1 not_normal 2 (50%)    1 (33.33%) 2 (100%)
3 skin      2 normal     2 (66.67%) 1 (50%)    3 (75%) 
4 skin      2 not_normal 1 (33.33%) 1 (50%)    1 (25%)

How to calculate category frequencies for all variables for each group separately with absolute and percentage values in R

1 Answers