[Babase] Grp on RANKS

Daphne A. Onderdonk babase@www.eco.princeton.edu
Tue, 28 Sep 2004 11:23:59 -0400


Hello all,

Sorry for the delay in responding to this - I wasn't very clear on 
Karl's question, and I guess I was hoping for someone to provide some 
clarity.  But after looking at the Groups table, rereading the 
documentation, and talking to Jeanne, we think we understand the 
question.  Jeanne recalls that this system of groups and supergroups was 
put into place to deal with the fissions - animals could be placed into 
a temporary group (as far as Members is concerned?), and that if the 
group is not permanent, the group would be different from the supergroup.

As Susan pointed out, right now, the GID column and the Supergroup 
column are the same in Groups, so right now it doesn't really matter.  
But there could be cases in the future where they are different, so it 
could matter.  Also, in looking at the Groups table, we wondered why 
group 1.4 (Alto's split - contains some males whose records were unclear 
during the split) is listed as permanent with a supergroup of 1.4.  We 
thought this maybe should a case where it is not permanent and has a 
supergroup of 1.0.  Thoughts?

So the question is what does the Ranker program currently use, and what 
do we want it to use.  There are 2 parts to the Ranker program - the 
part that creates the matrices, and the part that updates the Ranks 
table.  When the program is run, it asks for a "group ID".  Karl, when 
it creates a matrix from this, does it take the supergroup of that group 
ID (if they are different), and put everybody in the matrix who's in 
that supergroup (according to Members), or does it just include those 
present in that group (sub-group)?  Which does Ranker use to put into 
the Ranks table when it updates Ranks?  I think, from your question and 
from the documentation of Ranks, it uses Supergroup.  But I don't think 
we would want it to use a Supergroup in the Ranks table if it's only 
including the sub-group members in the matrix.  And do we want to be 
able to rank individuals within a sub-group?

Jeanne would like to discuss this tomorrow in our conference call, so 
this will hopefully give you all something to start thinking about in 
the meantime.  We're on for tomorrow at 1:00, right?

Daphne

Karl O. Pinc wrote:

> The documentation says this has to be a supergroup
> (one of the Supergroup column values on GROUPS).
> I assume that the ranker program is checking this and
> that this is true, right?
>
> Karl <kop@meme.com>
> Free Software:  "You don't pay back, you pay forward."
>                  -- Robert A. Heinlein
> _______________________________________________
> Babase mailing list
> Babase@www.eco.princeton.edu
> http://www.eco.princeton.edu/mailman/listinfo/babase