November 8, 2012

And the winner is ....

I am thrilled to announce that the 1st place winner of the Big Data 2012 Hackathon was DBMusic.

 
 

After a day of intense hacking, teams had to present their projects and respond to judges questions. It was an amazing moment to see so many teams excited about Couchbase server and developing innovative ideas powered by Couchbase.



The DBMusic project converts NASDAQ quote data into dubstep style music. By using different music tones (a saw, base wobble, some secondary wobble and percussion), you can get a feeling of what is going on in the data. It’s an interesting project for us because Couchbase Server was used in this project. After the event, we decided to speak to the winning team to learn why Couchbase server was the best choice among several other database options - here’s why  :


  1. Stock market transactions occur at blazing high speeds - couchbase Server provided sub-millisecond read and write latency that was needed by DBMusic.

  2. By using Couchbase Server, DBMusic was able to deliver high throughput to feed into the music synthesizers.  

  3. Mapreduce in Couchbase Server 2.0 provided simple real-time analytics on the data.

 

Here is a sample view that was created using Couchbase :

 

function (doc, meta) { // produce a time-based linear sequence of keys

 var st_seconds = parseInt(parseInt(doc.start_time) / 1000);

 var st_milliseconds = parseInt(doc.start_time) - (st_seconds * 1000);

 var st_minutes = parseInt(st_seconds / 60 % 60);

 var st_hours = parseInt(st_seconds / 60 / 60);

 var st_seconds = st_seconds % 60;

 var sd_year = parseInt(doc.start_date.substr(0, 4));

 var sd_month = parseInt(doc.start_date.substr(4, 2));

 var sd_day = parseInt(doc.start_date.substr(6, 2));

 emit([sd_year, sd_month, sd_day, st_hours, st_minutes, st_seconds, st_milliseconds], parseFloat(doc.ask_price));

}

If you're interested in checking out the full project, visit the DBMusic github repository.

Finally, don’t forget to check out the awesome final output track here 

Enjoy !

 

 

 

 
 
 

 

Comments