SCOUG-SundialSIG Mailing List Archives

<< Previous Message << >> Next Message >>

Date:	Fri, 1 Jun 2001 16:47:32 PDT
From:	"Lynn H. Maxson" <lmaxson@pacbell.net >
Reply-To:	scoug-sundialsig@scoug.com
To:	< "scoug-sundialsig@scoug.com" > scoug-sundialsig@scoug.com >
Subject:	SCOUG-SundialSIG: June meeting

Content Type: text/plain

Peter Skye wrote:

"Q: What is the "content" and the "context" for the following
three data
records?

"The quick brown fox jumped over the lazy dog."
"It's crackers to slip a rozzer the dropsy in snide."
"Baile, baile la bamba, yo no necesito un poco de gracias."

As Humphrey Bogart would say, "Here's looking at you". The fact
that you included each line of text within quotes indicates that
it is a single attribute (of text type), thus a single field (in a
record, a single column (in a row). It would frankly be true even
without the quotes if these existed as separate lines of text. If
they did not, if they constituted a paragraph say, then their
entirely is treated as a single unit, a single attribute. If they
were a section, a chapter, a book, or a library, in whatever
manner they were stored as a single whole, they still represent
only a single attribute (field, column, etc.).

So if you have an input "stream" of data, you have to make some
serious decisions when it comes to storing it. If you preprocess
it prior to storage, i.e. decompose and selectively extract
textual components of your definition, then I will guarantee you
that whatever unstructured nature they may have had in input has
been lost when you "tuck" them into your database or spreadsheet.
We impose structure on data, even on unstructured data.

"It's easier to teach the macro/program how to process the
different "types" of data files we receive than it is to create
some "master plan" (a data format) into which all data files must
fit."

Tsk, tsk, Master Peter, read what you have written. Your only
saving grace here is the word "all" which implies that you have
multiple "master plans", i.e. data formats, into which you fit
"all" your data (at least the extracted portions). You have to
because any system that you use, any application you have written
or any written by others, does not allow any other method than a
structuring one.

If the multiple is more convenient and easier to use (and
implement), then it just saves you from the embarrassment of
having to constantly change the "master file" due to the dynamics
of your data environment. I will guarantee that you will use some
translation process, if only in your head, that will allow you to
deal with multiple formats at a time. Otherwise the data in
whatever format is useless, if it cannot serve your purpose.

I would think in investment data capture when dealing with
multiple "fixed" sources and live feeds that you do as much as
possible to only extract those portions of text data that suits
your needs plus organize those extractions (impose a format) that
renders them more suitably to your use. Perhaps I err. Maybe you
do not follow the Principle of Least Effort.

=====================================================

To unsubscribe from this list, send an email message
to "steward@scoug.com". In the body of the message,
put the command "unsubscribe scoug-sundialsig".

For problems, contact the list owner at
"rollin@scoug.com".

=====================================================

Return to [ 01 | June | 2001 ]

SCOUG, Warp Expo West, and Warpfest are trademarks of the Southern California OS/2 User Group. OS/2, Workplace Shell, and IBM are registered trademarks of International Business Machines Corporation. All other trademarks remain the property of their respective owners.