* Before running the following DO file, rename variables "5th percentile", "10th percentile", "95th percentile", "97.5th percentile" and "99th percentile" 
* according to Stata's naming conventions: the first character of a name must be a letter or an underscore (it is recommended, however, that a variable name 
* starts with a letter since all of Stata's built-in variables begin with an underscore)

encode exposurehierarchyl1, gen(foodex_l1)
egen sumy1 = total(mean), by(foodex_l1)
generate perc = (mean/ sumy1)*100
gsort food -mean 
replace perc = -perc
bysort food (perc) : gen cum_perc = sum(perc)
replace cum_perc = -cum_perc
drop sumy1 foodex_l1
replace perc = -perc