[Babase] Conversion Errors - Round 2

Catherine Markham babase@www.eco.princeton.edu
Tue, 30 Aug 2005 10:19:40 -0400


Hi everyone,

We're on to round 2 of the conversion errors! Leah's notes have been cut 
and pasted into the email below - my notes and questions are inserted 
into that text.

Thanks,
Catherine

Leah Gerber wrote:

> Conversion errors – Take 2
> August 25, 2005
> 
> 1.    ERROR: WEE: 1993-09-09: Statdate of dead individuals must be >= 
> 1993-12-30

Corrected.

> 2.    ERROR: NEL: 1994-07-29: Statdate of live individuals must be = 
> 1994-02-23
> 
>     Can’t figure out why this statdate is as it is. In the Census table 
> NEL’s last entry is 30/12/93 in grp 2.0. In the members table last entry 
> is 24/02/94 in grp 2.0 (I). This corresponds to what I found in the 
> field notebook. NEL was censused on 23/02/94 in grp 2 and was absent 
> 25/02/94. There is one note in the subgroup notes but this does not help 
> place him after this time. No record in demography notes (except a 
> record of his absence from 4/03/94), other group notes or other group 
> censuses.

Corrected. I see NEL's last confirmed presence in CENSUS (status = "C" 
and cen = "T") placing him in group 2.0 on 23 February 1994, not 30 
December 1993.  Since Leah's searches didn't turn up any demography 
notes placing him in a group past his last census date, I changed the 
statdate for NEL changed from 29 July 1994 to 23 February 1994 .

> 3.    ERROR: MCC: 1990-01-29: Statdate of live individuals must be = 
> 1993-09-30
> 
>     One note in other group censuses placing MCC in grp 7 on 05/01/90. 
> Also seen in group 7  on 29/01/90  (Alto’s other group notes). I am 
> assuming statdate is correct.
> 
> NOTE:  I have noticed a bunch of individuals who have this end date of 
> 30/09/90 in members. A bunch of the weird males do and it is wrong. 
> Seems to be an artificial date that may be an artifact of some change to 
> the database. This may be something that we should look into in the 
> future. Some will be cleared up when I fix the males but there may be 
> some others. I keep running into this. A lot of the Group 99’s also have 
> this end date in members. Weird!

Leah and I talked about this one on Friday - correction involves
deleting many rows from MEMBERS and CENSUS.  I've put this one on hold 
until we get Karl's input for how best to correct.  See listserv message 
(subject name of "Conversion Errors - Deleting MEMBERS and CENSUS rows").

> 4.  The following individuals do not show up in census or members but 
> their statdate is not their birthdate.
> 
> ERROR:UT4: Birth 1997-09-18: Statdate 1997-10-08: Statdate of live 
> individuals, without census rows (absences excepted), must be the birth 
> date
> 
> ERROR:NZ5: Birth 1997-09-19: Statdate 1997-10-08: Statdate of live 
> individuals, without census rows (absences excepted), must be the birth 
> date
> 
> ERROR:CS1: Birth 1993-06-20: Statdate 1994-11-25: Statdate of live 
> individuals, without census rows (absences excepted), must be the birth 
> date
> 
> ERROR:DM1: Birth 1997-10-22: Statdate 1998-10-14: Statdate of live 
> individuals, without census rows (absences excepted), must be the birth 
> date
> 
> ERROR:D13: Birth 1998-02-20 Statdate 1998-08-22: Statdate of live 
> individuals, without census rows (absences excepted), must be the birth 
> date
> 
> ERROR:MN1: Birth 1997-12-30: Statdate 1998-02-17: Statdate of live 
> individuals, without census rows (absences excepted), must be the birth 
> date
> 
> ERROR:MT5: Birth 1998-02-26: Statdate 1998-11-19: Statdate of live 
> individuals, without census rows (absences excepted), must be the birth 
> date
> 
> The matgrps and sexes are as follows:
> 
> NZ5    F    3.2
> UT4    U    3.2
> CS1    F    1.3
> DM1    F    3.1
> D13    U    3.1
> MN1    M    3.2
> MT5    U    3.2
> 

This problem is coming up because we have entered some data for
Joy's, Nzige's, and Proton's groups at the time when each group stopped
being a regular study group (around late 1997 for Joy's and Nzige's and
early 1994 for Proton's).  Specifically, the births and statdates
(defined as last confirmed presence) of these seven indvidiuals have
been entered.   However, group census info for the time interval between 
birth and statdate has not been entered in Babase - yet.  One possible 
solution to this problem is adding the group census data to the database 
- I can see pros and cons of this.

Can we add this issue to the list of discussion topics for our next 
conference call?