Different results while working with stata in windows and mac

[Samir Thapa]

Hi All,

I've recently started working on Stata (13.1) on a mac system, and I'm getting different results compared to a windows system; with the following syntax with which I'm trying to obtain the number of technology constructed each fiscal year

ge recodate = date(created_on, "DMY", 2020)

format recodate %tdDD/NN/CCYY

sort recodate

gen fiscal_year3 = year(recodate)

replace fiscal_year3 = fiscal_year3 - 1 if recodate < mdy(7, 17, recodate)

sort fiscal_year3
by fiscal_year3: count if recodate != 0

I was working on the ms windows earlier, and extracted the excel file from windows to mac, and now when I run stata I get two different results as below, note that the differences are only from 1999/2000

Fiscal Year Total (Mac) Total (Windows)

1994/95 5115 5115
1995/96 7157 7157
1996/97 8387 8387
1997/98 9869 9869
1998/99 11052 11052
1999/00 13854 13265
2000/01 17257 14121
2001/02 15857 15134
2002/03 16021 20354
2003/04 11960 11352
2004/05 17195 17797
2005/06 16754 16115
2006/07 17089 16265
2007/08 14729 14239
2008/09 19522 15049
2009/10 22597 26806
2010/11 19151 19091
2011/12 18048 15551
2012/13 21856 26751
2013/14 37548 28446
2014/15 23824 30606
2015/16 14866 3844
2016/17 332 13674
2017/18
2018/19 6288 6021
2019/20 2617 2650
Total 368,945 368,711

Has it got to do with the date system with excel in mac and windows, which I dont think so, as when I convert the date systems it just adds four years to the existing dates, which gives me totally different results.

Thank you.

[William Lisowski]

I will note that you have a mistake in your code.

ge recodate = date(created_on, "DMY", 2020)
format recodate %tdDD/NN/CCYY
sort recodate
gen fiscal_year3 = year(recodate)
replace fiscal_year3 = fiscal_year3 - 1 if recodate < mdy(7, 17, recodate)

The mdy() function takes as arguments a month number, a day number, and a year. In your second line of code, recodate is a date value, not a year.

[Samir Thapa]

Thank you so much. Problem solved. Much appreciate.