working with dates to generate age

[Denise Vella]

Hi i tried to generate an age variable by subtracting (datecycle2) from dob = date of birth
I end up getting long numbers that don't make sense

I suspect it's because perhaps by dob is not in the correct format?
However i can't spot the mistake

It automatically got generated in the right way why I imported excel file into stata - it was imported as a %td format as a integer rather than float
meanwhile for the datecycle2 - this is a converted stata date which i did with the -date-command, which is correct.

Not sure about the dob.

What am I doing wrong?

* Example generated by -dataex-. For more info, type help dataex
clear
input int dob float(datecyle2 age)
 7034 22719 15685
 8189 23261 15072
 5548 23222 17674
10323 23232 12909
 9352 23248 13896
 2403 22384 19981
-3435 22475 25910
 1050 23247 22197
end
format %td dob
format %td datecyle2

[Denise Vella]

It's just occurred to me the numbers for age are being given in days , just need to /365
thanks anyway!

[Clyde Schechter]

Well, every four years there is a leap year, so if the people in question are adults, you will be somewhat off if you divide by 365. Dividing by 365.25 will be a bit better. Better still, if you are using a current version of Stata is to use the -age()- function which takes this into account exactly and even lets you specify what you want to consider as the birthday in non-leap years of somebody born on 29 Feb in a leap year. -help age()-. Or if you want the age including the fractional part, it's the -age_frac()- function.