[Babase] Statdate rules
Karl O. Pinc
babase@www.eco.princeton.edu
Tue, 14 Sep 2004 16:37:35 -0500
After conversation with Daphnie (and Steph) there's
some database rules for Statdate.
When there's a censused location for the individual
and the individual is still alive then the statdate
is the most recent census date. Note that demog
rows always have census rows to go with them
and act like C census rows.
Daphnie was a leery of this until she remembered
that she's never had to manually adjust a statdate.
Should such an adjustment be necessary a demography
note can always be made. (Such notes can be entered
with the updemog equalivent, or by hand.)
An individual can have any statdate >= birth when
they've not been censused.
A dead individual can have any statdate >= the last
census date.
The (new) last step in the conversion is to (try to)
compare the old foxpro database tables with the new
postgresql tables. We'll do sql queres and
compare the text. This seems like a good idea
and easy to do, we'll see if it realy is.
However, because the database now automatically sets
the statdate for live individuals, somebody
(Steph?) should check the statdates of all live
individuals to see that they correspond with the most
recent (non-absent) census rows. Unless there are no
(non-absent) census rows in which case it dosen't
matter. Be sure to check those (8?) individuals
that currently have no statdate but
will have one assigned before they get converted.
Karl <kop@meme.com>
Free Software: "You don't pay back, you pay forward."
-- Robert A. Heinlein