orngTxt
demo
home | what we do | background | about us | scenarios | contact | demo | benchmarks
NYSE TAQ

This demo is comprised of 987 million records from the New York Stock Exchange's Trades and Quotes (TAQ) database. Each record is a stock market quote, with ten fields: - ticker symbol, date, exchange, time, bid price, bid quantity, offer price, offer quantity, mode and MMID.   Most of the data are numeric, but ticker symbol, exchange, and MMID are text.

We received the data on more than 100 CDs, with no indexes; the Quotes portion of the database was compressed by the NYSE to approximately 36 gigabytes. After we copied the CDs to a hard drive, it took about a week of computing time to extract the Quotes data.

We have compressed the data to just over 4.5 gigabytes - small enough to fit on a single DVD.  We've indexed all ten fields, and added three indexes: month, day of month, and hour; the indexes total less than 4.7 gigabytes, so also fit on a single DVD.  You can expect to find around ten million records a second on a typical notebook computer, if you copy the indexes to the hard drive; this will improve as we continue to enhance the software

There are a few gaps in the data (notably, the NYSE didn't give us January data), and this is not an official NYSE release, but a demonstration of our compression and search technology.  Check back for revised software (bug fixes and performance improvements), as well as additional demos.

Please email us at feedback@xpace.net with bug reports, comments, and suggestions, or to order the full demo on two DVDs.

Instructions
  1. Create a folder (or two), and copy the DVDs, the one million record subset, or the 25 million record subset to it/them. 
  2. Unzip the NYSE demo software to a convenient place - perhaps the folder you created in
    step 1.
  3. Run the demo (NYSEDemo.exe).  Click "File", "Select Database Files", and pick the data and index files you copied in step 1.
  4. Type your query in the rectangles provided (e.g. "IBM" in the "Symbol" rectangle; "<=5" in "Bid"), and hit "Enter" or click on "Search."
Limitations
  1. Our list control allows only 100 million rows, so, although we can find any subset of the database, that is the maximum number of results we can display.  We are looking; suggestions welcome.
  2. We currently support only AND, OR, TO, <, >, >=,<=. 
orngLn