[Babase] Update trouble - Demog notes cannot have no note (nonulls allowed)?

Jeanne Altmann altj at princeton.edu
Tue Sep 7 11:44:21 EDT 2010


Excellent, thanks. And welcome back,
jeanne

-----Original Message-----
From: babase-bounces at eeblistserv.princeton.edu
[mailto:babase-bounces at eeblistserv.princeton.edu] On Behalf Of Karl O.
Pinc
Sent: Tuesday, September 07, 2010 11:35 AM
To: The Baboon Database Project
Subject: Re: [Babase] Update trouble - Demog notes cannot have no note
(nonulls allowed)?

Hi,

It's taken a while to get through my email backlog.

It has also taken me a while to work through the implications
of having a NULL DEMOG.Comment.  My analysis is below and
is probably worth reading.

In the end I came to the conclusion that everyone has made
the right decision and a NULL DEMOG.Comment is the right
way to go.  The change has been made to the database
and the documentation will be updated very shortly.

----------------<snip>-------------

Allowing the demography note to be null is not the right way
to go.  Instead of doing that we should remove the DEMOG
table entirely.  But I don't think that's the right approach
either.

I'm not sure exactly what the problem is that you're having,
or more precisely what exactly it is that you want to accomplish,
so let me list the various issues that I see.  I'm going to think
out loud as I go.

There are two possible sides to the problem.  A data entry problem
and a record keeping problem.  Record keeping is paramount,
although of course the data must go in, so let me start with
record keeping.  Once we know what we want to do with that
we can have further discussion regarding data entry.

There are a couple of possibilities.  First, you have no regular census
but do have a demography note, with a useless remark.  This is exactly
the sort of case where you would use the 'M' (manual) CENSUS.Status
value -- creating a manual census entry.  The 'M' Status was used
prior to 2001, I think by Daphine in an attempt to force interpolation
to do the right thing when we discovered that our early stabs
at interpolation were not working.  However the Status has not
been used since so, if it's really necessary to record that the
entry is the result of a demography note it could be used to mean
'census entry created as a result of a demography note'.  (In any
case I've a feeling that all those pre-2001 "M" statuses don't belong
and should be removed.  I could be wrong in this.)

This brings us to the heart of the issue, whether it's necessary
to record that the source of the CENSUS row is a demography note.
If it's not necessary to note that the source
of the CENSUS-row-that-does-not-appear-on-the-census-sheet is 
a demography note then 'M' works perfectly well. Alternately,
again, if there's no other reason for manually fussing with CENSUS
then the 'M' rows must be there because of demography notes.
If there are multiple sources of census-like data, demography notes,
something else, etc., that would cause you to make changes to CENSUS
then we could either create separate codes for each, which we would
want to do if it was important to query on this information, or
we could have a comment column that indicates the information's
source.  In the latter case, that's pretty much exactly what
the DEMOG.Comment column is -- and you'd enter in that column
something like "A demography note records that so-and-so
is in such-and-such group".  Exactly the comment which
started this thread, the comment that you want to get rid of!

:-)

(FWIW, I've a dim recollection of having had a discussion
like this one before.)

The second possibility is that there is already a CENSUS row
for the individual and there is a demography note placing
the individual in the group.  In this case I think you do
want to keep track of the existence of the demography note
so that if the record of the census gets changed or otherwise
goes away you need to keep a row on CENSUS because the
demography note exists.  Humm....  The right way to do
this is to allow NULL demography notes.

On 09/03/2010 04:35:55 PM, Niki Learn wrote:
> Yes, mostly it is with other group censuses.  There may also be
> occasional
> instances where a baboon is marked present using a demography note
> during an
> incomplete census.  
> 
> Thanks - Karl, please fix at your earliest convenience.
> 
> Niki
> 
> -----Original Message-----
> From: babase-bounces at eeblistserv.Princeton.EDU
> [mailto:babase-bounces at eeblistserv.Princeton.EDU] On Behalf Of Susan
> Alberts
> Sent: Friday, September 03, 2010 5:31 PM
> To: The Baboon Database Project
> Subject: Re: [Babase] Update trouble - Demog notes cannot have no 
> note
> (no
> nulls allowed)?
> 
> I don't completely understand what situations produce a demog note  
> that has a group and an ID number but no explanation-I presume that  
> this arises with other groups censuses? And sometimes with regular  
> group censuses?
> 
> If I'm understanding correctly then I agree with Jeanne; go with 1.
> 
> Susan
> On Sep 3, 2010, at 5:06 PM, Niki Learn wrote:
> 
> > Jeanne thinks 1 is best.  (2 is clunky and 3 is bad)
> >
> > If that's okay by Susan as well, Karl please remove the not-null  
> > constraint on the comment column in demog.  Then the upload can  
> > commence.
> >
> > Thanks!
> > Niki
> >
> > From: babase-bounces at eeblistserv.Princeton.EDU
> [mailto:babase-bounces at eeblistserv.Princeton.EDU 
> > ] On Behalf Of Niki Learn
> > Sent: Friday, September 03, 2010 4:26 PM
> > To: 'The Baboon Database Project'
> > Subject: [Babase] Update trouble - Demog notes cannot have no note  
> > (no nulls allowed)?
> >
> > I am working on the upload in test and have just come to the  
> > demography notes portion.  Demography notes also include census  
> > entries and other groups notes entries on nonstudy groups, where  
> > often the "note" would just consist of "so-and-so is present in
> such- 
> > and-such group".
> >
> > At the joint lab meeting we talked extensively about some revisions 
> 
> > to how we enter demography notes and what actually counts as a  
> > demography note.  One of the things we discussed was that these  
> > notes should be as un-wordy as possible and that the ones that just 
> 
> > say things like "so-and-so is present in such-and-such group", 
> which
>  
> > is already captured in the census record the demography note  
> > creates, that we didn't have to have any words because they were  
> > redundant.  So I implemented that for this upload and, just on the  
> > first group, got this error 9 times:
> >
> > CAUTION -- This error may not be real; prior row(s) of the uploaded 
> 
> > file were rejected:
> > Line 17: ERROR: null value in column "comment" violates not-null  
> > constraint
> >
> > So, the note can't be blank.  I see three possible options
> >
> > 1)      Remove the rule saying that the comment column (the actual  
> > note) cannot be null.  (This seems simplest.)
> > 2)      Go back to "so-and-so is present in such-and-such group".   
> > (Perhaps preferred  for some reason?)
> > 3)      Or I could enter those other groups entries the way we 
> enter
>  
> > census data - it just might have a lot of "N" entries in the file I 
> 
> > upload (but babase reads "N" as "don't create a record for this  
> > animal on this date" so it wouldn't create extra records or 
> anything
>  
> > like that).  The downside there being that if there were actual  
> > demography notes to go with any of them I would need a census 
> record
>  
> > and a demography note but it comes out the same in babase.  Also,  
> > entering them as demog notes tells you the reference source (i.e.,  
> > is there an actual census sheet in the other groups binder or is it 
> 
> > an other groups note within one of the study group binders and 
> which
>  
> > study group is it with).
> >
> > Thoughts?
> >
> > Thanks,
> > Niki
> >
> >
> > _______________________________________________
> > Babase mailing list
> > Babase at www.eco.princeton.edu
> > http://www.eco.princeton.edu/mailman/listinfo/babase
> 
> ------------------------------------------------------------
> Susan Alberts, Professor of Biology, Duke University, Box 90338,  
> Durham NC 27708. Phone 919-660-7272, FAX 919-660-7293
> 
> 
> 
> 
> 
> _______________________________________________
> Babase mailing list
> Babase at www.eco.princeton.edu
> http://www.eco.princeton.edu/mailman/listinfo/babase
> 
> _______________________________________________
> Babase mailing list
> Babase at www.eco.princeton.edu
> http://www.eco.princeton.edu/mailman/listinfo/babase
> 
> 




Karl <kop at meme.com>
Free Software:  "You don't pay back, you pay forward."
                 -- Robert A. Heinlein


_______________________________________________
Babase mailing list
Babase at www.eco.princeton.edu
http://www.eco.princeton.edu/mailman/listinfo/babase



More information about the Babase mailing list