[Babase] New Ranker login and misuse groomings

Lacey Maryott lacey.maryott at duke.edu
Tue Apr 1 14:21:45 EDT 2008


Tabby,

   Susan says you are brilliant :)   Nice Catch! we are so glad you 
caught that. Seems like a very easy fix. At first, Susan thought it odd 
that that could possibly explain why we were seei ng that happen with 
males, but we are pretty sure it was occuring mostly in subadult and 
juvenile males who do occasionally groom each other, so it seems it is 
possible that this could be the case... and obviously, even if acts 
aren't restricted, we wouldn't expect to see M&C's show up between two 
males or two females, so it definitely seems just to be a criteria 
restriction! :)

Hooray for obvious problems and not cryptic ones!

Cheers,
Lacey

kfenn wrote:
> As a follow-up to my earlier email regarding Ranker...
>
> I have successfully logged on and run the new program here at 
> Princeton (thanks for your help, Lacey!)
>
> Just for the my records (because my brain won't remember in 6 months)...
> the VPN connection requires one's papio username and login whereas the 
> Ranker program requires the Babase 2.0 (PhpPgAdmin) login.
>
> Lacey (earlier) and I (today) have discovered that the new Ranker 
> creates agonistic interaction within the matrices that DO NOT exist in 
> the actor_actees database.  I investigated this further....the source 
> of the extra data can be explained if one adds the grooming data to 
> the agonistic interactions.
> For example: new Ranker creates a matrix that shows ABB and EAG 
> having  4 interactions in Dec 06.  My old Ranker matrix showed only 1 
> interaction.  The actor_actees table shows ABB DS EAG 1 time, but ABB 
> G EAG 3 times.  (This pattern holds for all Omo dec 06 ALF data)  I 
> think the code for new Ranker must only do counts of the times that 
> ABB is an actor without discriminating between the 'act' codes.
> Now we need someone who can alter the code to make the distinction 
> because we can't use new Ranker as it is.   I won't pretend I've know 
> how to change the code, but the fix seems pretty straightforward.
>
>
> Tabby
>

-- 
Lacey Maryott
Alberts Lab
Department of Biology
Duke University
ph: 919-660-7306
fax: 919-660-7293
Lacey.Maryott at duke.edu 



More information about the Babase mailing list