Complicated data formation - r

Question

So I am trying to make a separate dataset that combines the yearly absence percentage and additionally binary variable of those with 10% or more total absence a year.
The absencePercentage should be calculated bycalculating total unauthorised and authorised absence divided by total possible sessions in all three terms.
Another thing is VioFlag. If the person has been flagged for Vio in at least one of the term, they should be flagged as VioFlagEver.
So the original data is like this:
ID PossibleSessions Term year unauthorisedAbsence authorisedAbsence VioFlag
0110 46 Sum 2014 0 1 0
0110 116 Win 2014 1 8 1
0110 56 Spr 2014 0 5 0
0110 44 Sum 2015 21 9 0
0110 120 Win 2015 2 2 0
0110 58 Spr 2015 10 1 0
So for ID 0110, he was absent for 15 sessions (0+1+1+8+0+5=15) out of possible 218 sessions (46+116+56=218). This means the absence percentage in 2014 for ID 0110 is 6.88%. He will not be the frequent absentee that year. But because in 2015, his absent rate was 20.27%, he will be a frequent absentee.
For ID 0110, He will be VioFlagEver for 2014 for not for 2015.
The new dataset I want to create is this.
ID year absencePercentage FrenquentAbsentee VioFlagEver
0110 2014 6.88 0 1
0110 2015 20.27 1 0
Please note that there are many IDs and year 2014 to 2018.
Thank you for your help!

Answer 1

You can try this:
library(tidyverse)
df %>% group_by(ID, year) %>%
summarize(absensepercentage = ((sum(unauthorisedAbsence) + sum(authorisedAbsence)) / sum(PossibleSessions))*100,
violflagever = if_else(sum(VioFlag) > 0, 1, 0),
frequentabsentee = if_else(absensepercentage > 10, 1, 0))

Answer 2

You can use tidyverse (dplyr) group_by and summarize to achieve this
library(tidyverse)
read.table(textConnection("ID PossibleSessions Term year unauthorisedAbsence authorisedAbsence VioFlag
0110 46 Sum 2014 0 1 0
0110 116 Win 2014 1 8 1
0110 56 Spr 2014 0 5 0
0110 44 Sum 2015 21 9 0
0110 120 Win 2015 2 2 0
0110 58 Spr 2015 10 1 0"),
header = T) %>%
as_tibble() -> df
df %>%
mutate(totalAbscence = unauthorisedAbsence+authorisedAbsence) %>%
group_by(ID, year) %>%
summarise(possibleAbscence = PossibleSessions %>% sum(),
totalAbscence = totalAbscence %>% sum(),
VioFlagEver = VioFlag %>% sum()) %>%
mutate(absencePercentage = (totalAbscence/possibleAbscence)*100,
FrenquentAbsentee = if_else(absencePercentage > 10, 1,0),
VioFlagEver = if_else(VioFlagEver > 0, 1, 0))
#> `summarise()` regrouping output by 'ID' (override with `.groups` argument)
#> # A tibble: 2 x 7
#> # Groups: ID [1]
#> ID year possibleAbscence totalAbscence VioFlagEver absencePercenta…
#> <int> <int> <int> <int> <dbl> <dbl>
#> 1 110 2014 218 15 1 6.88
#> 2 110 2015 222 45 0 20.3
#> # … with 1 more variable: FrenquentAbsentee <dbl>
Created on 2021-01-27 by the reprex package (v0.3.0)