How can I read dates with both 2-digit and 4-digit year? | Stata FAQ

Let’s read in some dates in string format.

clear
input str10 date_str
"4/16/2005"
"3/29/11"
"8/15/2009"
"5/14/07"
"5/3/08"
"10/10/2010"
end

list, clean

         date_str  
  1.    4/16/2005  
  2.      3/29/11  
  3.    8/15/2009  
  4.      5/14/07  
  5.       5/3/08  
  6.   10/10/2010

As you can see there is a mixture of 2-digit and 4-digit years. Data entry issues like this are more common than you might suspect. So, we will use the date function along with the %td format.

generate date=date(date_str,"MDY")
(3 missing values generated)

format %td date 

list, clean

         date_str        date  
  1.    4/16/2005   16apr2005  
  2.      3/29/11           .  
  3.    8/15/2009   15aug2009  
  4.      5/14/07           .  
  5.       5/3/08           .  
  6.   10/10/2010   10oct2010

All three of the dates with 4-digit years came out fine. We could use some string function or regular expression to insert the value 20 in front of the 2-digit years. However, a simper way is to use the date function with the mask “MD20Y“.

replace date=date(date_str,"MD20Y")
(3 real changes made)

list, clean

         date_str        date  
  1.    4/16/2005   16apr2005  
  2.      3/29/11   29mar2011  
  3.    8/15/2009   15aug2009  
  4.      5/14/07   14may2007  
  5.       5/3/08   03may2008  
  6.   10/10/2010   10oct2010

That worked just fine, but what if you had dates that go back to the previous century? Check out these dates.

clear
input str10 date_str
"4/16/2005"
"3/29/11"
"8/15/2009"
"5/14/07"
"5/3/08"
"10/10/2010"
"7/13/99"
"2/19/1997" 
end

list, clean

         date_str  
  1.    4/16/2005  
  2.      3/29/11  
  3.    8/15/2009  
  4.      5/14/07  
  5.       5/3/08  
  6.   10/10/2010  
  7.      7/13/99  
  8.    2/19/1997

We will start out as before.

generate date=date(date_str,"MDY")
(4 missing values generated)

format %td date 

list, clean

         date_str        date  
  1.    4/16/2005   16apr2005  
  2.      3/29/11           .  
  3.    8/15/2009   15aug2009  
  4.      5/14/07           .  
  5.       5/3/08           .  
  6.   10/10/2010   10oct2010  
  7.      7/13/99           .  
  8.    2/19/1997   19feb1997

Next, instead of one replace, we will use two. We will use the mask “MD20Y” if the last two values of date_str are less than “50”. When the last two values are greater than “50” we will use the mask “MD19Y”.

replace date=date(date_str,"MD20Y") if substr(date_str,-2,2)<"50" 
(3 real changes made) 

replace date=date(date_str,"MD19Y") if substr(date_str,-2,2)>"50"
(1 real change made)

list, clean

         date_str        date  
  1.    4/16/2005   16apr2005  
  2.      3/29/11   29mar2011  
  3.    8/15/2009   15aug2009  
  4.      5/14/07   14may2007  
  5.       5/3/08   03may2008  
  6.   10/10/2010   10oct2010  
  7.      7/13/99   13jul1999  
  8.    2/19/1997   19feb1997