/* open cleaned data generated in data_cleaning.do */
clear
set memory 80m
set more off
/* log using c:\lotteries\1997\output\elk_1997.log, replace */
use c:\lotteries\1997\intermediate\elk_res_1997_2.dta

/* assign id numbers and identify a unique huntcode for each huntcode */
drop choice_2-state resident-purchase age 
generate obsno=group(_N)
generate unique=0 if (lottery[_n]==lottery[_n+1]) 
replace unique=1 if unique==.
drop if obsno>18708
save c:\lotteries\1997\intermediate\elk_res_1997_3.dta, replace
clear

/* random draws of hunt codes; currently rifle only draws */
/* observations 1 to 1000 */
#delimit ;
for num 1/1000: use "c:\lotteries\1997\intermediate\elk_res_1997_3.dta" \
drop if unique==0 & obsno~=X \
drop if choice_1[_n-1]==choice_1[_n] \
sample 6.5 if (obsno<X | obsno>X) \ 
generate id=X \
generate choice= _n \
replace choice=0 if choice<10 \
replace choice=1 if choice>0 \
save "c:\lotteries\1997\intermediate\alt_X.dta", replace \ clear ;
#delimit cr

use c:\lotteries\1997\intermediate\alt_1.dta
for num 2/1000: append using "c:\lotteries\1997\intermediate\alt_X.dta"
for num 1/1000: erase "c:\lotteries\1997\intermediate\alt_X.dta"
generate chunks=group(1000)
drop unique orig_zip
save c:\lotteries\1997\intermediate\sorted_draws0.dta, replace
clear

/* observations 1001 to 2000 */
#delimit ;
for num 1001/2000: use "c:\lotteries\1997\intermediate\elk_res_1997_3.dta" \
drop if unique==0 & obsno~=X \
drop if choice_1[_n-1]==choice_1[_n] \
sample 6.5 if (obsno<X | obsno>X) \ 
generate id=X \
generate choice= _n \
replace choice=0 if choice<10 \
replace choice=1 if choice>0 \
save "c:\lotteries\1997\intermediate\alt_X.dta", replace \ clear ;
#delimit cr

use c:\lotteries\1997\intermediate\alt_1001.dta
for num 1002/2000: append using "c:\lotteries\1997\intermediate\alt_X.dta"
for num 1001/2000: erase "c:\lotteries\1997\intermediate\alt_X.dta"
generate chunks=group(1000)
drop unique orig_zip
save c:\lotteries\1997\intermediate\sorted_draws1.dta, replace
clear

/* observations 2001 to 3000 */
#delimit ;
for num 2001/3000: use "c:\lotteries\1997\intermediate\elk_res_1997_3.dta" \
drop if unique==0 & obsno~=X \
drop if choice_1[_n-1]==choice_1[_n] \
sample 6.5 if (obsno<X | obsno>X) \ 
generate id=X \
generate choice= _n \
replace choice=0 if choice<10 \
replace choice=1 if choice>0 \
save "c:\lotteries\1997\intermediate\alt_X.dta", replace \ clear ;
#delimit cr

use c:\lotteries\1997\intermediate\alt_2001.dta
for num 2002/3000: append using "c:\lotteries\1997\intermediate\alt_X.dta"
for num 2001/3000: erase "c:\lotteries\1997\intermediate\alt_X.dta"
generate chunks=group(1000)
drop unique orig_zip
save c:\lotteries\1997\intermediate\sorted_draws2.dta, replace
clear

/* observations 3001 to 4000 */
#delimit ;
for num 3001/4000: use "c:\lotteries\1997\intermediate\elk_res_1997_3.dta" \
drop if unique==0 & obsno~=X \
drop if choice_1[_n-1]==choice_1[_n] \
sample 6.5 if (obsno<X | obsno>X) \ 
generate id=X \
generate choice= _n \
replace choice=0 if choice<10 \
replace choice=1 if choice>0 \
save "c:\lotteries\1997\intermediate\alt_X.dta", replace \ clear ;
#delimit cr

use c:\lotteries\1997\intermediate\alt_3001.dta
for num 3002/4000: append using "c:\lotteries\1997\intermediate\alt_X.dta"
for num 3001/4000: erase "c:\lotteries\1997\intermediate\alt_X.dta"
generate chunks=group(1000)
drop unique orig_zip
save c:\lotteries\1997\intermediate\sorted_draws3.dta, replace
clear

/* observations 4001 to 5000 */
#delimit ;
for num 4001/5000: use "c:\lotteries\1997\intermediate\elk_res_1997_3.dta" \
drop if unique==0 & obsno~=X \
drop if choice_1[_n-1]==choice_1[_n] \
sample 6.5 if (obsno<X | obsno>X) \ 
generate id=X \
generate choice= _n \
replace choice=0 if choice<10 \
replace choice=1 if choice>0 \
save "c:\lotteries\1997\intermediate\alt_X.dta", replace \ clear ;
#delimit cr

use c:\lotteries\1997\intermediate\alt_4001.dta
for num 4002/5000: append using "c:\lotteries\1997\intermediate\alt_X.dta"
for num 4001/5000: erase "c:\lotteries\1997\intermediate\alt_X.dta"
generate chunks=group(1000)
drop unique orig_zip
save c:\lotteries\1997\intermediate\sorted_draws4.dta, replace
clear

/* observations 5001 to 6000 */
#delimit ;
for num 5001/6000: use "c:\lotteries\1997\intermediate\elk_res_1997_3.dta" \
drop if unique==0 & obsno~=X \
drop if choice_1[_n-1]==choice_1[_n] \
sample 6.5 if (obsno<X | obsno>X) \ 
generate id=X \
generate choice= _n \
replace choice=0 if choice<10 \
replace choice=1 if choice>0 \
save "c:\lotteries\1997\intermediate\alt_X.dta", replace \ clear ;
#delimit cr

use c:\lotteries\1997\intermediate\alt_5001.dta
for num 5002/6000: append using "c:\lotteries\1997\intermediate\alt_X.dta"
for num 5001/6000: erase "c:\lotteries\1997\intermediate\alt_X.dta"
generate chunks=group(1000)
drop unique orig_zip
save c:\lotteries\1997\intermediate\sorted_draws5.dta, replace
clear

/* observations 6001 to 7000 */
#delimit ;
for num 6001/7000: use "c:\lotteries\1997\intermediate\elk_res_1997_3.dta" \
drop if unique==0 & obsno~=X \
drop if choice_1[_n-1]==choice_1[_n] \
sample 6.5 if (obsno<X | obsno>X) \ 
generate id=X \
generate choice= _n \
replace choice=0 if choice<10 \
replace choice=1 if choice>0 \
save "c:\lotteries\1997\intermediate\alt_X.dta", replace \ clear ;
#delimit cr

use c:\lotteries\1997\intermediate\alt_6001.dta
for num 6002/7000: append using "c:\lotteries\1997\intermediate\alt_X.dta"
for num 6001/7000: erase "c:\lotteries\1997\intermediate\alt_X.dta"
generate chunks=group(1000)
drop unique orig_zip
save c:\lotteries\1997\intermediate\sorted_draws6.dta, replace
clear

/* observations 7001 to 8000 */
#delimit ;
for num 7001/8000: use "c:\lotteries\1997\intermediate\elk_res_1997_3.dta" \
drop if unique==0 & obsno~=X \
drop if choice_1[_n-1]==choice_1[_n] \
sample 6.5 if (obsno<X | obsno>X) \ 
generate id=X \
generate choice= _n \
replace choice=0 if choice<10 \
replace choice=1 if choice>0 \
save "c:\lotteries\1997\intermediate\alt_X.dta", replace \ clear ;
#delimit cr

use c:\lotteries\1997\intermediate\alt_7001.dta
for num 7002/8000: append using "c:\lotteries\1997\intermediate\alt_X.dta"
for num 7001/8000: erase "c:\lotteries\1997\intermediate\alt_X.dta"
generate chunks=group(1000)
drop unique orig_zip
save c:\lotteries\1997\intermediate\sorted_draws7.dta, replace
clear

/* observations 8001 to 9000 */
#delimit ;
for num 8001/9000: use "c:\lotteries\1997\intermediate\elk_res_1997_3.dta" \
drop if unique==0 & obsno~=X \
drop if choice_1[_n-1]==choice_1[_n] \
sample 6.5 if (obsno<X | obsno>X) \ 
generate id=X \
generate choice= _n \
replace choice=0 if choice<10 \
replace choice=1 if choice>0 \
save "c:\lotteries\1997\intermediate\alt_X.dta", replace \ clear ;
#delimit cr

use c:\lotteries\1997\intermediate\alt_8001.dta
for num 8002/9000: append using "c:\lotteries\1997\intermediate\alt_X.dta"
for num 8001/9000: erase "c:\lotteries\1997\intermediate\alt_X.dta"
generate chunks=group(1000)
drop unique orig_zip
save c:\lotteries\1997\intermediate\sorted_draws8.dta, replace
clear

/* observations 9001 to 10000 */
#delimit ;
for num 9001/10000: use "c:\lotteries\1997\intermediate\elk_res_1997_3.dta" \
drop if unique==0 & obsno~=X \
drop if choice_1[_n-1]==choice_1[_n] \
sample 6.5 if (obsno<X | obsno>X) \ 
generate id=X \
generate choice= _n \
replace choice=0 if choice<10 \
replace choice=1 if choice>0 \
save "c:\lotteries\1997\intermediate\alt_X.dta", replace \ clear ;
#delimit cr

use c:\lotteries\1997\intermediate\alt_9001.dta
for num 9002/10000: append using "c:\lotteries\1997\intermediate\alt_X.dta"
for num 9001/10000: erase "c:\lotteries\1997\intermediate\alt_X.dta"
generate chunks=group(1000)
drop unique orig_zip
save c:\lotteries\1997\intermediate\sorted_draws9.dta, replace
clear

/* observations 10001 to 11000 */
#delimit ;
for num 10001/11000: use "c:\lotteries\1997\intermediate\elk_res_1997_3.dta" \
drop if unique==0 & obsno~=X \
drop if choice_1[_n-1]==choice_1[_n] \
sample 6.5 if (obsno<X | obsno>X) \ 
generate id=X \
generate choice= _n \
replace choice=0 if choice<10 \
replace choice=1 if choice>0 \
save "c:\lotteries\1997\intermediate\alt_X.dta", replace \ clear ;
#delimit cr

use c:\lotteries\1997\intermediate\alt_10001.dta
for num 10002/11000: append using "c:\lotteries\1997\intermediate\alt_X.dta"
for num 10001/11000: erase "c:\lotteries\1997\intermediate\alt_X.dta"
generate chunks=group(1000)
drop unique orig_zip
save c:\lotteries\1997\intermediate\sorted_draws10.dta, replace
clear

/* observations 11001 to 12000 */
#delimit ;
for num 11001/12000: use "c:\lotteries\1997\intermediate\elk_res_1997_3.dta" \
drop if unique==0 & obsno~=X \
drop if choice_1[_n-1]==choice_1[_n] \
sample 6.5 if (obsno<X | obsno>X) \ 
generate id=X \
generate choice= _n \
replace choice=0 if choice<10 \
replace choice=1 if choice>0 \
save "c:\lotteries\1997\intermediate\alt_X.dta", replace \ clear ;
#delimit cr

use c:\lotteries\1997\intermediate\alt_11001.dta
for num 11002/12000: append using "c:\lotteries\1997\intermediate\alt_X.dta"
for num 11001/12000: erase "c:\lotteries\1997\intermediate\alt_X.dta"
generate chunks=group(1000)
drop unique orig_zip
save c:\lotteries\1997\intermediate\sorted_draws11.dta, replace
clear

/* observations 12001 to 13000 */
#delimit ;
for num 12001/13000: use "c:\lotteries\1997\intermediate\elk_res_1997_3.dta" \
drop if unique==0 & obsno~=X \
drop if choice_1[_n-1]==choice_1[_n] \
sample 6.5 if (obsno<X | obsno>X) \ 
generate id=X \
generate choice= _n \
replace choice=0 if choice<10 \
replace choice=1 if choice>0 \
save "c:\lotteries\1997\intermediate\alt_X.dta", replace \ clear ;
#delimit cr

use c:\lotteries\1997\intermediate\alt_12001.dta
for num 12002/13000: append using "c:\lotteries\1997\intermediate\alt_X.dta"
for num 12001/13000: erase "c:\lotteries\1997\intermediate\alt_X.dta"
generate chunks=group(1000)
drop unique orig_zip
save c:\lotteries\1997\intermediate\sorted_draws12.dta, replace
clear

/* observations 13001 to 14000 */
#delimit ;
for num 13001/14000: use "c:\lotteries\1997\intermediate\elk_res_1997_3.dta" \
drop if unique==0 & obsno~=X \
drop if choice_1[_n-1]==choice_1[_n] \
sample 6.5 if (obsno<X | obsno>X) \ 
generate id=X \
generate choice= _n \
replace choice=0 if choice<10 \
replace choice=1 if choice>0 \
save "c:\lotteries\1997\intermediate\alt_X.dta", replace \ clear ;
#delimit cr

use c:\lotteries\1997\intermediate\alt_13001.dta
for num 13002/14000: append using "c:\lotteries\1997\intermediate\alt_X.dta"
for num 13001/14000: erase "c:\lotteries\1997\intermediate\alt_X.dta"
generate chunks=group(1000)
drop unique orig_zip
save c:\lotteries\1997\intermediate\sorted_draws13.dta, replace
clear

/* observations 14001 to 15000 */
#delimit ;
for num 14001/15000: use "c:\lotteries\1997\intermediate\elk_res_1997_3.dta" \
drop if unique==0 & obsno~=X \
drop if choice_1[_n-1]==choice_1[_n] \
sample 6.5 if (obsno<X | obsno>X) \ 
generate id=X \
generate choice= _n \
replace choice=0 if choice<10 \
replace choice=1 if choice>0 \
save "c:\lotteries\1997\intermediate\alt_X.dta", replace \ clear ;
#delimit cr

use c:\lotteries\1997\intermediate\alt_14001.dta
for num 14002/15000: append using "c:\lotteries\1997\intermediate\alt_X.dta"
for num 14001/15000: erase "c:\lotteries\1997\intermediate\alt_X.dta"
generate chunks=group(1000)
drop unique orig_zip
save c:\lotteries\1997\intermediate\sorted_draws14.dta, replace
clear

/* observations 15001 to 16000 */
#delimit ;
for num 15001/16000: use "c:\lotteries\1997\intermediate\elk_res_1997_3.dta" \
drop if unique==0 & obsno~=X \
drop if choice_1[_n-1]==choice_1[_n] \
sample 6.5 if (obsno<X | obsno>X) \ 
generate id=X \
generate choice= _n \
replace choice=0 if choice<10 \
replace choice=1 if choice>0 \
save "c:\lotteries\1997\intermediate\alt_X.dta", replace \ clear ;
#delimit cr

use c:\lotteries\1997\intermediate\alt_15001.dta
for num 15002/16000: append using "c:\lotteries\1997\intermediate\alt_X.dta"
for num 15001/16000: erase "c:\lotteries\1997\intermediate\alt_X.dta"
generate chunks=group(1000)
drop unique orig_zip
save c:\lotteries\1997\intermediate\sorted_draws15.dta, replace
clear

/* observations 16001 to 17000 */
#delimit ;
for num 16001/17000: use "c:\lotteries\1997\intermediate\elk_res_1997_3.dta" \
drop if unique==0 & obsno~=X \
drop if choice_1[_n-1]==choice_1[_n] \
sample 6.5 if (obsno<X | obsno>X) \ 
generate id=X \
generate choice= _n \
replace choice=0 if choice<10 \
replace choice=1 if choice>0 \
save "c:\lotteries\1997\intermediate\alt_X.dta", replace \ clear ;
#delimit cr

use c:\lotteries\1997\intermediate\alt_16001.dta
for num 16002/17000: append using "c:\lotteries\1997\intermediate\alt_X.dta"
for num 16001/17000: erase "c:\lotteries\1997\intermediate\alt_X.dta"
generate chunks=group(1000)
drop unique orig_zip
save c:\lotteries\1997\intermediate\sorted_draws16.dta, replace
clear

/* observations 17001 to 18000 */
#delimit ;
for num 17001/18000: use "c:\lotteries\1997\intermediate\elk_res_1997_3.dta" \
drop if unique==0 & obsno~=X \
drop if choice_1[_n-1]==choice_1[_n] \
sample 6.5 if (obsno<X | obsno>X) \ 
generate id=X \
generate choice= _n \
replace choice=0 if choice<10 \
replace choice=1 if choice>0 \
save "c:\lotteries\1997\intermediate\alt_X.dta", replace \ clear ;
#delimit cr

use c:\lotteries\1997\intermediate\alt_17001.dta
for num 17002/18000: append using "c:\lotteries\1997\intermediate\alt_X.dta"
for num 17001/18000: erase "c:\lotteries\1997\intermediate\alt_X.dta"
generate chunks=group(1000)
drop unique orig_zip
save c:\lotteries\1997\intermediate\sorted_draws17.dta, replace
clear

/* observations 18001 to 18708 */
#delimit ;
for num 18001/18708: use "c:\lotteries\1997\intermediate\elk_res_1997_3.dta" \
drop if unique==0 & obsno~=X \
drop if choice_1[_n-1]==choice_1[_n] \
sample 6.5 if (obsno<X | obsno>X) \ 
generate id=X \
generate choice= _n \
replace choice=0 if choice<10 \
replace choice=1 if choice>0 \
save "c:\lotteries\1997\intermediate\alt_X.dta", replace \ clear ;
#delimit cr

/* this merges each of the random draws for the 1000-individual subsets */ 
use c:\lotteries\1997\intermediate\alt_18001.dta
for num 18002/18708: append using "c:\lotteries\1997\intermediate\alt_X.dta"
for num 18001/18708: erase "c:\lotteries\1997\intermediate\alt_X.dta"
generate chunks=group(1000)
drop unique orig_zip
save c:\lotteries\1997\intermediate\sorted_draws18.dta, replace
clear

use c:\lotteries\1997\intermediate\sorted_draws0.dta
for num 1/18: append using "c:\lotteries\1997\intermediate\sorted_drawsX.dta"
drop id
generate id=group(18708)
save c:\lotteries\1997\intermediate\sorted_draws.dta, replace
for num 0/18: erase "c:\lotteries\1997\intermediate\sorted_drawsX.dta"

/* merge origin zipcodes with the randomly drawn hunts */
use c:\lotteries\1997\intermediate\elk_res_1997_3.dta
keep obsno orig_zip 
rename obsno id
sort id 
save c:\lotteries\1997\intermediate\elk_res_1997_4.dta, replace
clear

use c:\lotteries\1997\intermediate\sorted_draws.dta
sort id
merge id using c:\lotteries\1997\intermediate\elk_res_1997_4.dta, nokeep
drop _merge chunks resobsno
save c:\lotteries\1997\intermediate\sorted_draws.dta, replace
clear

use c:\lotteries\shared\elk_miles.dta
rename huntcode choice_1
sort choice_1 orig_zip
save c:\lotteries\1997\intermediate\elk_miles_2.dta, replace
clear

use c:\lotteries\1997\intermediate\sorted_draws.dta
sort choice_1 orig_zip
merge choice_1 orig_zip using c:\lotteries\1997\intermediate\elk_miles_2.dta, nokeep
drop _merge
rename miles miles_1
sort id obsno
save c:\lotteries\1997\intermediate\sorted_draws.dta, replace
clear

/* this is code for merging miles with second choice */
/*
use c:\lotteries\shared\elk_miles.dta
rename huntcode choice_2
sort choice_2 orig_zip
save c:\lotteries\1997\intermediate\elk_miles_2.dta, replace
clear

use c:\lotteries\1997\intermediate\sorted_draws.dta
replace choice_2="E-1-164" if choice_2=="E-I-164"
replace choice_2="- -" if choice_1=="E-3-407" & name_f=="GARY" & name_l=="NUNN"
sort choice_2 orig_zip
merge choice_2 orig_zip using c:\lotteries\1997\intermediate\elk_miles_2.dta, nokeep
rename miles miles_2
drop _merge
sort choice_1 name_l
*/

/* this merges the hunt characteristics with the random draws */
use c:\lotteries\1997\data\hunts_97.dta
sort huntcode
rename huntcode choice_1
save c:\lotteries\1997\intermediate\hunts_1997_2.dta, replace

use c:\lotteries\1997\intermediate\sorted_draws.dta
sort choice_1
merge choice_1 using c:\lotteries\1997\intermediate\hunts_1997_2.dta, nokeep
sort id obsno
drop _merge
save c:\lotteries\1997\intermediate\sorted_draws, replace
clear

/* this merges applicant characteristics (gender, age) with the random draws */
use c:\lotteries\1997\intermediate\elk_res_1997_2.dta
drop choice_2-resident phone-purchase
generate obsno=group(_N)
drop if obsno>18708
keep obsno gender age 
rename obsno id
sort id
save c:\lotteries\1997\intermediate\elk_res_1997_5.dta, replace
clear

use c:\lotteries\1997\intermediate\sorted_draws.dta
sort id
merge id using c:\lotteries\1997\intermediate\elk_res_1997_5.dta, nokeep
drop _merge
save c:\lotteries\1997\intermediate\sorted_draws.dta, replace
clear

/* this merges the drawn/purchased data with the random draws */
use c:\lotteries\1997\intermediate\elk_res_1997_2.dta
keep choice_1 drawn purchase
generate obsno=group(_N)
drop if obsno>18708
keep obsno choice_1 drawn purchase
rename obsno id
sort id choice_1
save c:\lotteries\1997\intermediate\elk_res_1997_5.dta, replace
clear

use c:\lotteries\1997\intermediate\sorted_draws.dta
sort id choice_1
merge id choice_1 using c:\lotteries\1997\intermediate\elk_res_1997_5.dta, nokeep

generate outcome=1 if choice_1==drawn
replace outcome=0 if outcome==.
generate outcome2=1 if (purchase~=. & purchase~=0) & choice_1~=drawn
replace outcome2=0 if outcome2==.
generate male=1 if gender=="M"
replace male=0 if male==.
generate constant=1
generate prob_prc=licenses/apps95
replace prob_prc=1 if prob_prc>1
generate prob_tru=licenses/apps96
replace prob_tru=1 if prob_tru>1
generate harvest=charv/chunts
replace harvest=1 if harvest>1
replace miles=925 if miles==.
generate age_sqd=age*age
drop gender _merge drawn purchase

gsort id -choice
drop orig_zip marquez mba rifle bow muzzle handicap
save c:\lotteries\1997\intermediate\10_sorted_draws.dta, replace
drop choice_1
outsheet using c:\lotteries\1997\data\10_sorted_draws.txt, nonames