[Babase] Issues with newcycles and newcycstats
Catherine Markham
babase@www.eco.princeton.edu
Wed, 16 Feb 2005 11:52:56 -0500
Hi Karl,
Laurence and I wanted to give you a bit more information on the funky
things going on with the new cycles tables . . .
It seems the heart of the trouble is a problem in newcycles - it looks
like mdates were calculated only for resumes after a pregnancy, as
opposed to all cycles other than resumes. One of the ramifications in
newcycstats arising from the incorrect newcycles table is that there are
gaps from ddate to next tdate if an mdate is not present in newcycles.
Perhaps the program is reading lack of an mdate as an indication of
pregnancy?
If it would help in any way, Laurence has put together a correct cycles
table where she calculated mdates where applicable (all cycles but
resumes) and when possible (previous cycle has a ddate). In addition,
she has corrected calculated mdate values in this table to adjust for
any mdates > tdates. We had hoped that we could use this correct table
to replace your newcycles table and then rerun the newmakerep program
and generate a correct newcycstats. Unfortunately, this didn't work at
all. We realized the newmakerep program was working to recreate the
newcycles table, along with newcycstats and newrepstats. Don't worry -
we didn't change anything on my computer in playing around with this.
We just did everything with a copy of the files on Laurence's computer,
so all of the tables and programs on my computer are just as you
originally had them.
Thanks for your help and let us know if there's anything we can do on
our end.
Catherine and Laurence