Re: st: Problem with the time

[Nick Cox <njcoxstata@gmail.com>]

There are languages in which dates are distinct data types, but in
Stata the idea is
to hold a date in an appropriate numeric data type and to provide a
rich set of date formats and conversion functions. It would be pretty
bizarre and inefficient to insist that yearly dates must be converted
to milliseconds just because some uses require ms.

Actually, it is entirely possible to work outside Stata's date system
even when using dates.
For example, with daily meteorological data, it is customary to move
back and forth between daily, monthly and yearly resolution and having
separate day -- month -- year variables is the best way to do that.
Having a single date variable that is mdy(month, day, year) is
essential for some purposes but would frustrate many others.

Nick

On Wed, Apr 13, 2011 at 5:54 PM, Guido Lüchters
<guidoluechters@gmail.com> wrote:
> Dear Nick
> yes this is completely clear
> but sometimes i am thinking: what if there was one internal date
> representation within stata
>
> On the other hand: the stata form is most flexible in respect to
> import date values
> like study-time in any unit
> so pros/cons
> But this time i was just not prepared
>

*
*   For searches and help try:
*   http://www.stata.com/help.cgi?search
*   http://www.stata.com/support/statalist/faq
*   http://www.ats.ucla.edu/stat/stata/