[Babase] informal update on Babase corrections
kfenn
kfenn at princeton.edu
Wed Mar 14 16:02:50 EDT 2007
Hi Everyone,
I'm providing an informal update on my part of the error checking as
discussed in the conference call on Friday. I'm also including a little
background for babase users who weren't involved in the conference call
but want to know what this error checking means about the overall status
of getting Babase online.
The last conversion run produced only 1 rank error. All other known
errors have been fixed. (These known errors were based upon the rule
checking that Karl has programmed into the online system.) It also
produced a load of errors related to the INTERACT table which I believe
is related to a program bug that Karl found. (I'm not clear on the
specifics of this, so be sure to check with someone if you need to
immediately access the current INTERACT data.)
The current version of FoxPro data has been directly uploaded into PPA
in the database "babase". The results of the conversion run have been
loaded into the database "babase_test" in PPA. What does this mean
exactly? It means that the PPA 'babase' is a carbon copy of the Foxpro
'babase'....warts and all. So that 1 remaining rank error is still an
error in the PPA 'babase". However, in 'babase_test', the data ran
through the filter of the conversion program, so the one remaining rank
error probably prevented a tiny amount of rank data from loading
correctly.
You can currently use the PPA 'babase' and you will have ALMOST exactly
the same data that is available in the FoxPro version. The one
difference is the "CYCLES" table and the tables or views built from this
(repstats, cycstats, mtd_cycles). Some errors in this table could not
be detected by the error checks in the conversion. Therefore, Karl has
run a comparison to detect the differences between the FoxPro and PPA
Cycles tables. Why is there a difference? Because we now have a new
table called "CYCAPS" in PPA, the m, t, and d dates of a cycle are
sometimes grouped differently than in FoxPro. Sometimes this new
grouping fixes an error, and sometimes it creates one because CYCGAPS is
incorrect. It's difficult to say whether FoxPro or PPA has a 'more
correct' CYCLES table so use either table for now.
I have been reviewing Karl's output for the past 2 days. I expect to
begin making corrections to CYCGAPS tomorrow. It could take easily take
a week to make the necessary corrections as working with the CYCGAPS
table is tricky for me. Currently I have about 125 cycles that need to
be checked and potentially updated. I can update you next week about
the progress.
Finally, Leah and Lacey are in possession of the 'official' FoxPro
version of babase and are doing updates for 2006a. Karl will run
another conversion after they upload the data. There may be some errors
to correct, but I think we may be may be within striking distance of an
'error free' version of babase on PPA, current through Jun06, by roughly
the end of March. The 2006b demography update will occur after all this.
Please email corrections or clarifications if I didn't get something right.
Tabby
--
Tabby Fenn
Research Assistant
Dept of Ecology and Evolutionary Biology
401 Guyot Hall
Princeton University
Princeton, NJ 08544
609 258-6898 (Ph)
609 258-2712 (Fx)
More information about the Babase
mailing list