[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

RE: Data Accuracy



PureBytes Links

Trading Reference Links

Great study.

While the computers in the study were identical, the results don't 
necessarily predict the results of our studio audience.  The number of lost 
ticks will vary with a number of factors including

Number of symbols in portfolio
Number of symbols having real time data stored
Speed of the processor (A P5-200 will loose more ticks than a PII-233)
Amount of RAM
Speed of hardisk access and amount of space
Operating system (say NT vs. 95/98)
Serial port type

In other words, though the study machines showed a data loss of 23%, a 
faster machine with more RAM may only loose 1% because its able to 
distribute news data faster.  A similar machine with a slower processor and 
less RAM could loose even more than 23%.

The BMI news program may not suffer the same penalties.

-----Original Message-----
From:	James F. Mazzulla [SMTP:tagteam@xxxxxxx]
Sent:	Saturday, June 27, 1998 8:38 PM
To:	omega-list@xxxxxxxxxx
Subject:	Re: Data Accuracy

On Sat, 27 Jun 1998 Richard C. wrote:
>The computer which did not collect news data collected 23.8% more data
>(ticks) than the computer collecting news. This percentage was calculated 
on
>file size not the number of ticks.

I run News continually. Based upon a comparison with other
TS users and the CME website, I do miss some ticks. The
loss, however,  only appears to be around 1% (on average)
and I haven't been able to determine whether it's related to
running News or something else. Regardless, 23% is =huge=
and I simply can't imagine a loss of that magnitude being
related =exclusively= to collecting news.

Best regards,
Jim