How to make a frequency table by class [duplicate]

Question

This is the code I have set up so far :

library(dslabs)
library(dplyr)
library(lubridate)

data("reported_heights")

dat <- mutate(reported_heights, date_time = ymd_hms(time_stamp)) %>%
  filter(date_time >= make_date(2016, 01, 25) & date_time < make_date(2016, 02, 1)) %>%
  mutate(type = ifelse(day(date_time) == 25 & hour(date_time) == 8 & between(minute(date_time), 15, 30), "inclass","online")) %>%
  select(sex, type, time_stamp)

y <- factor(dat$sex, c("Female", "Male"))
x <- dat$type

counter <- count(dat, sex,type)

It creates for me a tbl_df that looks like this, link below :

      sex | type    | n 
1  Female | inclass | 26
2  Male   | inclass | 13
3  Female | online  | 42
4  Male   | online  | 69

I am asking if you can help me with a code that will calculate the proportion of each sex in each type of class.

I have been trying to create a new table using the x characters "inclass" and "online" as columns with a proportion column added and then the y factors "male" and "female" would be the rows. I have been trying to do this using pull() and prop.table() but I am a total newbie and it would mean the world to me if you beautiful experts can help me. I have been going through answers for hours now and maybe the answer is already out there so please excuse that I can't seem to find it.... Thank you so much.

What is the proportion of the sexes(male&female) in each type of class(inclass&online)?

It's possible to calculate this by dividing the sex with the total number of students in a given type of class.

For example: There are 42 females studying online out of the total (42+69)=111. Answer: In the online class 38% are females.

How can we do this in R ?

Answer 1

Using prop.table():

prop.table(table(y, x), 2)
#        x
#y          inclass    online
#  Female 0.6666667 0.3783784
#  Male   0.3333333 0.6216216

Answer 2

You may use table(),

my.table <- with(dat, table(sex, type))
my.table
#         type
# sex      inclass online
#   Female      26     42
#   Male        13     69

and apply() a function on the result.

res <- apply(my.table, 2, function(x) x/sum(x)*100) 
res
#         type
# sex       inclass   online
#   Female 66.66667 37.83784
#   Male   33.33333 62.16216

To get a nicer output you could round() then and add %.

res2 <- as.data.frame(unclass(round(res, 1)))
res2[] <- lapply(res2, paste0, "%")
res2
#        inclass online
# Female   66.7%  37.8%
# Male     33.3%  62.2%

Answer 3

To get the proportion in each class, we can use ave in base R

df$prop <- with(df, n/ave(n, type, FUN = sum)) * 100 
df
#     sex    type  n     prop
#1 Female inclass 26 66.66667
#2   Male inclass 13 33.33333
#3 Female  online 42 37.83784
#4   Male  online 69 62.16216

The same can be achieved with dplyr

library(dplyr)
df %>% group_by(type) %>% mutate(prop = n/sum(n) * 100)

and data.table

library(data.table)
setDT(df)[, prop := n/sum(n) * 100, by = type]

data

df <- structure(list(sex = structure(c(1L, 2L, 1L, 2L), .Label = c("Female", 
"Male"), class = "factor"), type = structure(c(1L, 1L, 2L, 2L
), .Label = c("inclass", "online"), class = "factor"), n = c(26L, 
13L, 42L, 69L)), class = "data.frame", row.names = c(NA, -4L))