[Babase] Robert's comments on Psion header info
Susan Alberts
alberts at duke.edu
Sat Jan 7 14:06:42 EST 2006
Karl, Catherine, Leah,
Below are Robert's comments on information in the Psion headers,
relevant to the questions we talked about yesterday.
>X-Sieve: CMU Sieve 2.2
>From: Robert Zimmerman <rzman at duke.edu>
>Subject: Re: DISREGARD LAST EMAIL, read this one
>Date: Sat, 7 Jan 2006 11:34:45 -0500
>To: Susan Alberts <alberts at duke.edu>
>
>Here are a few thoughts about these obscure parts of the
>ptsamplr/psion scheme...
>
>The palmtop id is a single character (right now 1-4, I believe)
>that's used to distinguish the machines. It's main purpose is to
>keep the data separate when it's dumped and copied onto a
>laptop--the dump file names (restricted to 8 characters on the
>psion) are made from the date and the psion id character.
>
>The setup file id and program id are intended to help keep track of
>significant structural/functional changes in both the software and
>the sampling procedure. As I remember I put them in at first to make
>sure that the data I was generating to test the system didn't get
>mixed up with real data. The values used in these fields are not
>terribly systematic and longer than they need to be. Since the
>psions have come into use I think there should be maybe a dozen
>different values for either field (or fewer), and 3 or 4 that cover
>the bulk of the data.
>
>Of the two, the setup file id has the most relevance for
>managing/analyzing the data. It should point to a configuration file
>that's archived somewhere (they're on my computer, if nothing else)
>that specifies the format of each kind of data record, the valid
>codes in each field, etc. It's likely to be useful someday to sort
>out exactly what was or wasn't collected. Changes in this code also
>indicate significant changes in sampling methodology. I think there
>have also been some short transitional periods that correspond to a
>specific setup file id, when a new sampling procedure or software
>feature or sample type has been introduced but hasn't been quite
>worked out. The setup file id changes (after a week or so, probably)
>as sampling or software problems are worked out.
>
>The program id is probably less relevant, and changes in the program
>id and setup file id will generally track each other. It may be that
>in some of those transitional periods, as bugs are worked out, the
>setup file doesn't change but the program id does.
>
>I'm not sure that all of this has to be preserved in the database,
>though I would think that it should be. It can always be extracted
>from the original data, if need be. But it should fit naturally in
>the database and it has the potential to be pretty useful now and
>then.
>
>Susan, it might be a good thing to do this summer to just make a
>quick annotated list of the setup file id/program file id
>combinations over the years. It wouldn't be a very long list and
>could be a useful little reference in years to come. It might also
>be good to make sure that the old setup files are archived more or
>less systematically in a place where they can be found by whoever
>might come along trying to analyze the data.
>
>On Jan 5, 2006, at 3:20 PM, Susan Alberts wrote:
>
>
>**** **** **** **** **** **** **** **** **** ****
>* Robert Zimmerman
>* Duke University Dept. of Music
>* rzman at duke.edu * (919) 660-3303
--
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
Susan Alberts, Associate Professor
Department of Biology, Duke University, Box 90338, Durham NC 27708
phone 919-660-7272 fax 919-660-7293
More information about the Babase
mailing list