How can I count the total number of occurrences at time step t of an element?

Question

I am a beginner in R but I would like to write a piece code that requires some R and data science knowledge.

I have a data frame with the following structure; t1 denotes 10 minutes time periods and 1 defines measurements.

 t1 t2 t3 t4
  1  0  0  0
  1  1  1  1
  0  1  1  1
  0  1  1  1
  1  0  1  1

I would like to identify the duration and starting point for each measurement. For example, there are two 10 min measurements starting at t1 (row 1 and row 5) and there are two 30 minutes measurements that start at t2 (row 3and row 4).

Output:

 duration_minutes   t1 t2 t3 t4
      10            2   0  0  0
      20            1   0  1  0
      30            0   2  0  0
      40            1   0  0  0

Is there a way to convert the counts into percentages?

df<-structure(list(t1 = c(1, 1, 0, 0,1),
                   t2 = c(0, 1, 1, 1,0), t3 = c(0, 1, 1, 1,1), t4 = c(0, 1, 1,  1,1)), row.names = c(NA,5L), class = "data.frame") 

                                                                                  

Answer 1

I have solved it by first pivoting the data into the long format and then counting consecutive 1's to determine the duration of each measurement. Then I count how often each duration occurs per starting time and pivot back to the wider format to get the output that you described (I do not find a column t4 since it doesn't have any new measurements):

library(tidyr)
library(dplyr)

df %>% 
  mutate(rownr = 1:nrow(.)) %>% 
  pivot_longer(names_to = 'time', values_to = 'value', cols = 1:4) %>% 
  group_by(rownr, grp = cumsum(value == 0)) %>% 
  mutate(duration = 10 * cumsum(value)) %>% 
  filter(duration != 0) %>% 
  summarise(time = first(time), 
            duration = max(duration)) %>% 
  group_by(time, duration) %>% 
  count() %>% 
  pivot_wider(names_from = time, values_from = n, values_fill = 0) %>% 
  arrange(duration)