[1] 
* load plot-level data
    use             "$data/plot_dataset.dta", clear

* basic structure check
    describe
    sum             harvest_kg nitrogen_kg plot_area_GPS hh_id_merge season
	
[2]

* create yield per hectare
    gen             yield = harvest_kg / plot_area_GPS

* quick check
    sum             yield
    *** mean 22431, min -519.8, max 7.51e+08

* check for missing or zero `plot_area_GPS`
    count           if plot_area_GPS == 0 | plot_area_GPS == .
    *** 15,442 missing area

[3] 

* create combined id
    egen            hh_season_id = group(hh_id_merge season)
    lab var         hh_season_id "household-season identifier"

* spot check
    list            hh_id_merge season hh_season_id in 1/10
    *** confirm that hh_season_id is constant within each hh_id_merge-season combination