Create a grouping variable based on previous 2 years

[John MacDonald]

I have event date data where there are multiple evendates. I'd like to create a grouping variable based on the past two years starting in 2020. Given their are multiple events on the same date I can't tsset the data without first collapsing it.

ro_n = unique identifier

eventdate year ro_n weight
18jun2017 2017 2 6
18jun2017 2017 2 5
29jan2020 2020 2 3
29jan2020 2020 2 1
29jan2020 2020 2 14
29jan2020 2020 2 6
29jan2020 2020 2 2
29jan2020 2020 2 6
29jan2020 2020 2 9
29jan2020 2020 2 15
29jan2020 2020 2 4
26nov2022 2022 3 15
26nov2022 2022 3 10
26nov2022 2022 3 6
26nov2022 2022 3 14
26nov2022 2022 3 7
29oct2023 2023 4 2
29oct2023 2023 4 13

[John MacDonald]

Here is what I did but I'm sure someone has a more efficient method.

* Calculate the year difference from 2020
gen year_2023=year-2021
gen year_2022=year-2020
gen year_2021=year-2019
gen year_2020=year-2018

* Grouping variable based on the past two years starting from 2023
foreach var of varlist year_2023 year_2022 year_2021 year_2020 {
gen g`var'=0
replace g`var'=1 if `var'==0
replace g`var'=1 if `var'==1
}

[Nick Cox]

See also https://www.statalist.org/forums/for...uping-variable

I don't follow how these new variables can be useful, but the three commands inside the loop can be condensed to one:

gen g`var' = cond(inlist(`var', 0, 1), 1, 0)