[Babase] Auto mdates - for Jeanne and Karl

Catherine Markham amarkham at princeton.edu
Fri Jun 2 15:40:21 EDT 2006


Hi all,

I've been checking the auto mdates and things generally look really good 
so far.  For the most part, it seems the errors that I'm coming onto are 
due to mistakes in the original CYCLES table (so still nothing that's 
wrong with the new programming).

Just one question for now...

If you look at time interval between an automatically generated mdate 
and the following tdate of that same cycle, you get a range extending 
from 0 days (that part makes sense) all the way up to over 250 days 
(yikes - seems really long).

Jeanne, when you get a chance: I'm assuming 250 days is a red flag.  I 
can look into the records more carefully and try and figure out whether 
we missed coding a gap in monitoring or if the cycle data are otherwise 
incomplete.  But first I need to know above what interval length I 
should be concerned with.  Is, say, less than 60 days probably an 
acceptable interval between mdate and tdate but 61 or greater something 
to investigate?  70 days?  50 days?  I'm just pretty clueless as to what 
the "norm" might be.

On a slightly different note, I've done some other checking and it looks 
like the mdates are indeed being generated following the basic 
documentation rules.  I'll keep playing around with it, though.  Karl, 
I'll give you a head's up when I think I've exhausted all the checks I 
can think of.  So far, so good, though!!

Catherine


More information about the Babase mailing list