Find the count or frequency over rolling periods by group

Question

Here's my data:

Person<-c("Bob","Bob","Bob","Bob","Mary","Mary","Mary","Mary","Sue","Sue","Sue","Sue")
Year<-c(1980,1981,1982,1983,1980,1981,1982,1983,1980,1981,1982,1983)
Attendance<-c(1,1,0,0,1,1,1,0,0,0,1,0)
data<-as.data.frame(cbind(Person,Year,Attendance))
data

 Person Year Attendance
    Bob 1980          1
    Bob 1981          1
    Bob 1982          0
    Bob 1983          0
    Mary 1980         1
    Mary 1981         1
    Mary 1982         1
    Mary 1983         0
    Sue 1980          0
    Sue 1981          0
    Sue 1982          1
    Sue 1983          0

I would like to create a table using tidyverse/dplyr that shows the frequency of attendance for each person over rolling two-year periods. Like this:


Person  Two_year_period Frequency_attendance
Bob     1980-1981   100%
Bob     1981-1982   50%
Bob     1982-1983   0%
Mary    1980-1981   100%
Mary    1981-1982   100%
Mary    1982-1983   50%
Sue     1980-1981   0%
Sue     1981-1982   50%
Sue     1982-1983   50%

Guillermo Garcia · Accepted Answer · 2022-04-07T16:57:13.493

Here's a solution using dplyr:

data_final <- data %>% 
  group_by(Person) %>%
  mutate(Two_year_period  = paste(Year,lead(as.numeric(Year)), sep = "-")) %>% # Create new column with lead year name
  mutate(Attendance2 = lead(as.numeric(Attendance))) %>% # Create new column with lead attendance value
  drop_na(Attendance2) %>% # Drop NA
  mutate(Frequency_attendance = paste((as.double(Attendance) + as.double(Attendance2))*(100/2), "%")) %>% # Calculate Frequency_attendance
  select(-one_of(c('Year', 'Attendance', 'Attendance2'))) #Drop auxiliary columns
data_final

  Person Two_year_period Frequency_attendance
  <chr>  <chr>           <chr>               
1 Bob    1980-1981       100 %               
2 Bob    1981-1982       50 %                
3 Bob    1982-1983       0 %                 
4 Mary   1980-1981       100 %               
5 Mary   1981-1982       100 %               
6 Mary   1982-1983       50 %                
7 Sue    1980-1981       0 %                 
8 Sue    1981-1982       50 %                
9 Sue    1982-1983       50 %

This is excellent! I modified it for my actual dataset and calculated frequencies for rolling six-year-periods. Thank you so much, Guillermo! — Tish, Apr 07 '22 at 20:00
I'm very glad to hear that it helped! If you consider my answer is right, accepting as a right answer would be much appreciated. Best of luck to you @Tish! — Guillermo Garcia, Apr 07 '22 at 20:58

Find the count or frequency over rolling periods by group

1 Answers1