Currently, I am collecting the closing quotes by deleting "stock_data",
then accumulating quotes for a couple of hours after the close of the stock
exchange.
The resulting file is sorted with duplicate records suppressed, renamed to
"dat.YY.MM.DD" , compressed, and stored away for later processing.
Now, I am looking to analyze the data collected. While it is efficient to
collect the quotes into a single file per day, the analysis of a single stock
is more efficient if the data is rearrainged to be one stock per file.
We see, therefore that the data must be "cooked" before it can really be
crunched.
I would be most interested to hear if there is a "standard" for the format of
such data that should be used so that the data collected might be intermingled
with other data and processed by commonly available analysis programs.
Lacking such a standard at this point, let me suggest this format based upon
the information available in the quotes coming from X*Press.
YY.MM.DD:NAME:BID:ASK:HIGH:LOW:LAST:VOLUME
Where YY Year
MM Month
DD Day
NAME Symbol
BID Floating Point number
ASK Floating Point number
HIGH Floating Point number
LOW Floating Point number
LAST Floating Point number
VOLUME Integer
Example:
93.10.27:AAPL:31.50:31.75:32.25:29.75:31.75:123
Rationale:
Including both the date and the stock symbol in each record makes the data
portable between different files. All of the information is there in the one
record, and the name of the file(s) is not critical. I did drop the "change"
field from the data because that is simply calculated by comparing the
previous record. If information is not available, then the field is null -
that is if BID and ASK are not available in the original record, then
the cooked record becomes:
93.10.27:AAPL:::32.25:29.75:31.75:123
Comments? Help?
-- Richard Schaal Motorola Computer Group M/S AZ43 DW278 2900 South Diablo Way, Tempe, AZ 85252 Voice: (602)438-3519 Fax: (602)438-3836 E-Mail: rschaal@phx.mcd.mot.com "This is a great place to work.... there's a going away party every Friday!"