[Babase] Dupe Cenids in DEMOG
Catherine Markham
babase@www.eco.princeton.edu
Tue, 23 Aug 2005 15:43:23 -0400
Karl,
Leah forwarded the following conversion error to me:
> 2. DEMOG Cenid = 1673602, 1676301
>
> There are two entries for these cenid’s.
I looked into these and here's my best guess at what went wrong: the
"double" cenids reference two separate demography notes that match with
regard to individual baboon, group, and date - the only difference is
the reference group. For example, the "dupe" cenids for 1673602 both
place Icarus in Viola's group in 5 February 2001, but the reference for
one of these demography notes is Viola's group (with a note reading
"ICA Joins Viva's today, seen consorting with VET; last seen with
Nyayo's;") and one is Nyayo's group (with a note reading "ICA missing
from Nyayo's").
Also, in both cases there was an actual census done that day on the
study group that the demography note places the baboon in (CENSUS row
has status = C and cen = T) and in the other reference group (CENSUS row
has status = A and cen = T). For example, ICA was marked present on the
group census for Viola's group on 5 February 2001 and absent that day
for the group census for Nyayo's group. A "present" group placement for
him that day was essentially entered in triplicate (2 demography notes
plus 1 group census).
So these dupes could just be fluke, but I wanted to err on the side of
caution and give you a summary of what I was able to track down. My
question is whether or not this highlights something we should look into
with valdemog - possibly a complication that happens with redundant
demog note info? Or maybe this is something you have corrected in more
recent versions of the validation program for demography notes?
Thanks,
Catherine
P.S. In terms of the "short-term" fix and correcting the problem for
these 2 particular cenids, I think I'm all set: in each case, I'm going
to (1) combine the note info from the two separate demography notes into
a single note, (2) delete the duplicate demography note, and (3) run
reinterp on the baboon. I doubt the last step is necessary since both
demography notes place the individual in the same group, but I want to
make sure none of the changes to DEMOG inadvertently alter MEMBERS.
Please let me know if this doesn't sound like a good plan.