[Babase] Supergroup Issue
Lacey Maryott
lacey.maryott at duke.edu
Wed May 14 11:12:17 EDT 2008
Karl O. Pinc wrote:
>
> On 05/13/2008 10:19:55 PM, Lacey Maryott wrote:
>>
>>
>> Quoting "Karl O. Pinc" <kop at meme.com>:
>>
>>>
>>> On 05/13/2008 02:32:07 PM, Lacey Maryott wrote:
>>>> Hi Karl,
>>>>
>>>> I have just started testing the 07B update, and I decided to try
>>>> th potentially most complcated data first, and found my expectation
>>>> to be correct. For some reason I am getting an error of 'ERROR:
>>>> Cannot find supergroup of group 2.00 on 1979-11-01: group 2.00 has no
>>>> supergroup before 1980-07-01.'
>
> I'd say the problem is in the data. Why is Hooks not "permanent"
> before 1980-07-01 if you want to put in ranks that predate
> that date?
>
> 1: You can't be ranked in a temporary sub-group unless all individuals
> ranked are in the same supergroup.
>
> 2: A group has no supergroup before it's Permanent date if it's
> got no From group.
>
> So, you can't rank in a permanent group with no From group before
> it's Permanent date. I think we've used this to provide sanity
> checks for rank dates during the conversion from Foxpro. Maybe.
This makes sense
> (I know there were some rank date errors found.) This would
> imply that the 1980-11-01 date is an artificial value. I don't
> recall enough of the group's history to be sure. Perhaps this is
> a Jeanne question? My guess would be that you want to slide that
> date back to the date of the earliest valid ranking.
I spoke with Susan, and she said that sliding back the permanent date to
match rankings seems the best solution here. Since these rankings were
only recently reconstructed from the notebooks, it essentially changes
what the database previously 'knew' about the beginnings of this group.
The earliest I have complete ranks for this group is 1979-11. Should I
change this in the groups table, Karl?
>
> This is probably worth a "Tip" in the GROUPS documentation.
> Let me know what you think. Some suggested wording would be nice too.
>
> Historically we've had to fiddle with the dates in the Groups file
> when fission's occurred and such to get the rankings right.
> I thought that Babase 2.0 eliminated such fiddling. I could be wrong
> so please check that what I'm saying here makes sense.
>
> 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
>
--
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