[Babase] male dispersal query

Susan Alberts alberts at duke.edu
Mon May 17 21:20:09 EDT 2010


Thanks Karl, see below.

On May 14, 2010, at 5:33 PM, Karl O. Pinc wrote:

> Hi,
> Susan and I just spoke on the phone and we decided
> that two new interfaces (views or other view-like things
> that appear to be tables) are called for.
>
> A GROUPMOVES view.  This would be one row for
> each time an individual changed groups.  It would
> show the last day in the old group, the first day in
> the new group, the old group, the new group, and the
> sname.

This is what I had in mind for the MALEMOVES view. For the GROUPMOVES  
view I had in mind a table that had the following columns:

Source Group
Destination Group
Number of males that moved in that direction

It would produce a matrix with each cell representing the number of  
males that had moved from the source to the destination group.

However, upon reflection I see that it makes much more sense to just  
create a MALEMOVES view and create the matrix from the elements in the  
MALEMOVES view, because that way you'd be able to indicate time  
periods, etc.


>
> A MALEMOVES view.  This is the view that Susan describes below.
> However Susan had a caveat in that she expressed a desire to
> ignore "small" absences from a group, where "small" is variable
> and can be different for group 9 than for other groups.

Yes, this remains an issue - how do we ignore short-duration migration  
events (1-2 days) or events that involve unknown time periods breaking  
up residencies.
>
> One question that Susan and I didn't discuss are birth
> and statdate and how they would appear (or not) in the
> results.  If anyone has any comment please let me know.
> FWIW it'd be easiest if neither the group membership
> that starts at birth nor the group memebership that ends
> with (or after) the statdate appear.  That does not sound
> like a good feature though so most likely I'll have
> the "prior group" for birth and the "next group" for
> statdate be NULL.

Yes, this sounds right. We don't want to eliminate the first or last  
groups the male was in.
>
> I'll see about working on this.  I'll also read Niki's email
> more closely and see what we need to address there.
>
> FWIW these are all self-join issues, although the "small
> absence fuzz factor" makes things tricky and I may not use
> a view at all in this aspect of the implementation.  (Translation:
> You may not be able to write a sane query that does this.)

Yes, understood, thanks.

Hope you're feeling better.

Susan

>
> On 05/14/2010 02:16:40 PM, Susan Alberts wrote:
>> HI Karl,
>>
>> Here is a problem that has come up before and will again.
>>
>> We'd like to be able to extract data from members of the following
>> form, regarding male dispersal and group membership patterns
>>
>> Group, Date into that group, Date out of that group, Sname
>>
>> Where each male has an exhaustive set of rows that shows where he is
>> for all the chunks of his life. I don't know how to get this out of
>> members easily.
>>
>> Can you provide advise? Is this something we could set up as a view
>> sometime?
>
>
>
>
>
>
> 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

--------------------------------------------------------
Susan Alberts, Dept of Biology, Duke University, Box 90338, Durham NC  
27708, 919-660-7272 (Ph), 919-660-7293 (Fax)






More information about the Babase mailing list