R: Group by values in a column and count each value

Question

I have a simple data frame. The elements are individuals, the variables are activities, and the values are whether each individual has completed the activity.

df1 <- data.frame (student = c (1, 2, 3, 4, 5, 6),
               budget= c("in progress", "not started", "completed", "not started", "not started", "in progress"),
               resume = c ("not started", "completed", "completed", "not started", "completed", "in progress"),
               cover = c("completed", "not started", "not started", "not started", "in progress", "in progress"))

I want to create a table where the rows are "completed," "in progress," and "not started," the columns are the activities ("budget," "resume," and "cover"), and the values are the count for each.

I have tried using the "group_by" function.

dt1 <- df1 %>% 
  group_by (budget, resume, cover) %>% 
  summarise(freq = n())

But this seems to be counting a combination of values.

What I want in the end is a table that looks like

df2 <- data.frame (progress = c("completed", "in progress", "not started"),
                   budget = c(1, 2, 3),
                   resume = c(3, 1, 2),
                   cover = c(1, 2, 3))

Any and all feedback is appreciated. Thank you.

Answer 1

In tidyverse you may reshape the data to long format, count and reshape to wide format.

library(dplyr)
library(tidyr)

df1 %>%
  pivot_longer(cols = -student) %>%
  count(name, value) %>%
  pivot_wider(names_from = name, values_from = n)

However, I think it is easier in base R in this case -

sapply(df1[-1], table)

#            budget resume cover
#completed        1      3     1
#in progress      2      1     2
#not started      3      2     3