Calculating probability based upon share of years

[Muhammad Saleh]

I have firm level survey data (idstd) of almost 70,000 firms. Each firm is surveyed in (data_year) and has been mapped with total disbursements of WB funded projects two years before the survey (transaction_year) in the same state_province_admn1. My data looks like this

----------------------- copy starting from the next line -----------------------
[CODE]
* Example generated by -dataex-. To install: ssc install dataex
clear
input double idstd float data_year str43 state_province_admn1 str38 country float transaction_year double total_disbursements
300931 2001 "Tirana" "Albania" 1999 1904693960
301036 2001 "Tirana" "Albania" 1999 1904693960
300944 2001 "Tirana" "Albania" 1999 1904693960
300943 2001 "Tirana" "Albania" 1999 1904693960
300976 2001 "Tirana" "Albania" 1999 1904693960
301078 2001 "Tirana" "Albania" 1999 1904693960
301029 2001 "Tirana" "Albania" 1999 1904693960
300942 2001 "Tirana" "Albania" 1999 1904693960
301047 2001 "Tirana" "Albania" 1999 1904693960
300986 2001 "Tirana" "Albania" 1999 1904693960
301080 2001 "Tirana" "Albania" 1999 1904693960
301079 2001 "Tirana" "Albania" 1999 1904693960
301087 2001 "Tirana" "Albania" 1999 1904693960
300945 2001 "Tirana" "Albania" 1999 1904693960
300929 2001 "Tirana" "Albania" 1999 1904693960
301088 2001 "Tirana" "Albania" 1999 1904693960
300935 2001 "Tirana" "Albania" 1999 1904693960
300962 2001 "Tirana" "Albania" 1999 1904693960
300960 2001 "Tirana" "Albania" 1999 1904693960
301089 2001 "Tirana" "Albania" 1999 1904693960
301075 2001 "Tirana" "Albania" 1999 1904693960
300985 2001 "Tirana" "Albania" 1999 1904693960
301068 2001 "Tirana" "Albania" 1999 1904693960
301065 2001 "Tirana" "Albania" 1999 1904693960
301074 2001 "Tirana" "Albania" 1999 1904693960
301066 2001 "Tirana" "Albania" 1999 1904693960
301030 2001 "Tirana" "Albania" 1999 1904693960
301096 2001 "Tirana" "Albania" 1999 1904693960
301045 2001 "Tirana" "Albania" 1999 1904693960
301027 2001 "Tirana" "Albania" 1999 1904693960
301095 2001 "Tirana" "Albania" 1999 1904693960
300934 2001 "Tirana" "Albania" 1999 1904693960
300975 2001 "Tirana" "Albania" 1999 1904693960
301035 2001 "Tirana" "Albania" 1999 1904693960
300946 2001 "Tirana" "Albania" 1999 1904693960
300930 2001 "Tirana" "Albania" 1999 1904693960
300959 2001 "Tirana" "Albania" 1999 1904693960
301076 2001 "Tirana" "Albania" 1999 1904693960
300994 2001 "Tirana" "Albania" 1999 1904693960
300977 2001 "Tirana" "Albania" 1999 1904693960
301086 2001 "Tirana" "Albania" 1999 1904693960
301077 2001 "Tirana" "Albania" 1999 1904693960
301097 2001 "Tirana" "Albania" 1999 1904693960
301044 2001 "Tirana" "Albania" 1999 1904693960
300928 2001 "Tirana" "Albania" 1999 1904693960
300940 2001 "Tirana" "Albania" 1999 1904693960
300941 2001 "Tirana" "Albania" 1999 1904693960
300963 2001 "Tirana" "Albania" 1999 1904693960
301028 2001 "Tirana" "Albania" 1999 1904693960
306489 2001 "Absheron" "Azerbaijan" 1999 1063601040
306504 2001 "Absheron" "Azerbaijan" 1999 1063601040
306598 2001 "Absheron" "Azerbaijan" 1999 1063601040
306573 2001 "Absheron" "Azerbaijan" 1999 1063601040
306521 2001 "Absheron" "Azerbaijan" 1999 1063601040
306501 2001 "Absheron" "Azerbaijan" 1999 1063601040
306491 2001 "Absheron" "Azerbaijan" 1999 1063601040
306603 2001 "Absheron" "Azerbaijan" 1999 1063601040
306506 2001 "Absheron" "Azerbaijan" 1999 1063601040
306550 2001 "Absheron" "Azerbaijan" 1999 1063601040
306604 2001 "Absheron" "Azerbaijan" 1999 1063601040
306518 2001 "Absheron" "Azerbaijan" 1999 1063601040
306605 2001 "Absheron" "Azerbaijan" 1999 1063601040
306524 2001 "Absheron" "Azerbaijan" 1999 1063601040
306497 2001 "Absheron" "Azerbaijan" 1999 1063601040
306508 2001 "Absheron" "Azerbaijan" 1999 1063601040
306509 2001 "Absheron" "Azerbaijan" 1999 1063601040
306538 2001 "Absheron" "Azerbaijan" 1999 1063601040
306562 2001 "Absheron" "Azerbaijan" 1999 1063601040
306530 2001 "Absheron" "Azerbaijan" 1999 1063601040
306487 2001 "Absheron" "Azerbaijan" 1999 1063601040
306592 2001 "Absheron" "Azerbaijan" 1999 1063601040
306552 2001 "Absheron" "Azerbaijan" 1999 1063601040
306486 2001 "Absheron" "Azerbaijan" 1999 1063601040
306492 2001 "Absheron" "Azerbaijan" 1999 1063601040
306536 2001 "Absheron" "Azerbaijan" 1999 1063601040
306526 2001 "Absheron" "Azerbaijan" 1999 1063601040
306570 2001 "Absheron" "Azerbaijan" 1999 1063601040
306560 2001 "Absheron" "Azerbaijan" 1999 1063601040
306543 2001 "Absheron" "Azerbaijan" 1999 1063601040
306576 2001 "Absheron" "Azerbaijan" 1999 1063601040
306523 2001 "Absheron" "Azerbaijan" 1999 1063601040
306650 2001 "Absheron" "Azerbaijan" 1999 1063601040
306495 2001 "Absheron" "Azerbaijan" 1999 1063601040
306568 2001 "Absheron" "Azerbaijan" 1999 1063601040
306553 2001 "Absheron" "Azerbaijan" 1999 1063601040
306551 2001 "Absheron" "Azerbaijan" 1999 1063601040
306528 2001 "Absheron" "Azerbaijan" 1999 1063601040
306507 2001 "Absheron" "Azerbaijan" 1999 1063601040
306535 2001 "Absheron" "Azerbaijan" 1999 1063601040
306602 2001 "Absheron" "Azerbaijan" 1999 1063601040
306593 2001 "Absheron" "Azerbaijan" 1999 1063601040
306561 2001 "Absheron" "Azerbaijan" 1999 1063601040
306590 2001 "Absheron" "Azerbaijan" 1999 1063601040
306520 2001 "Absheron" "Azerbaijan" 1999 1063601040
306531 2001 "Absheron" "Azerbaijan" 1999 1063601040
306577 2001 "Absheron" "Azerbaijan" 1999 1063601040
306601 2001 "Absheron" "Azerbaijan" 1999 1063601040
306542 2001 "Absheron" "Azerbaijan" 1999 1063601040
306589 2001 "Absheron" "Azerbaijan" 1999 1063601040
306595 2001 "Absheron" "Azerbaijan" 1999 1063601040

I want to have a variable based upon the probability of each state_province_admn1 gettting the WB funding based upon the share of years each province gets funding out of total years. Like for transaction_year ranging (1999-2014) and if for state of Tirana who gets funding in 1999, 2002 and in 2013, its probability will be 3/16. Please advise how can I program? I tried egen with group function but its not prducing desired outcome.

[Clyde Schechter]

I don't understand. In "if for state of Tirana who gets funding in 1999, 2002 and in 2013," where do the 2002 and 2013 come from? Tirana only gets funding in 1999 (transaction_year). Or is it 2001 (data_year)?

Please clarify.

[Hemanshu Kumar]

Perhaps something like this?

egen byte tag_trans = tag(transaction_year)
egen num_trans_years = total(tag_trans)

egen byte tag_state_year = tag(state_province_admn1 country data_year)
bysort country state_province_admn1: egen num_state_years = total(tag_state_year)

gen wanted = num_state_years/num_trans_years
drop tag_trans tag_state_year num_state_years num_trans_years

[Muhammad Saleh]

Clyde Schechter Data is sorted by transaction_year and country. "data_year" is the year in which firm level survey was conducted. Variable of interest is transaction_year. transaction_year is the year of disbursements against WB funded projects. So in 1999 I have all the WB disbursements made in different states of different countries. Then in 2000 WB total disbursements made in different states of different countries and so on so forth upto 2014. So a state say Tirana can have disbursements in more than one years against WB funded projects. Hope I am able to clarify. Kindly let me know otherwise.
Hemanshu Kumar thanks sir I will check and let you know

[Muhammad Saleh]

Hemanshu Kumar Dear sir, thanks for your guidance it worked!
Clyde Schechter Thanks for your feedback. Just in case if you think some alternate way of doing the same please let me know as well