Shades: In-memory OLAP cubing, histograms, and more

#1

Post by d2fn » Mon, Jun 24, 2013, 9:36 PM UTC

Shades: In-memory OLAP cubing, histograms, and more
github.com

Re: Shades: In-memory OLAP cubing, histograms, and more

#2

Post by binarymax » Mon, Jun 24, 2013, 11:23 PM UTC

Nice project. I've messed around with cubes in the past. Curious what algorithm you used for the cube? I was mostly working with powersets and trees, and then experimented with dwarf petacubes. Do you have a limit on dimensions due to code, beyond what fits in memory?

Re: Shades: In-memory OLAP cubing, histograms, and more

#3

Post by d2fn » Tue, Jun 25, 2013, 12:00 AM UTC

binarymax wrote: post #2

Nice project. I've messed around with cubes in the past. Curious what algorithm you used for the cube? I was mostly working with powersets and trees, and then experimented with dwarf petacubes. Do you have a limit on dimensions due to code, beyond what fits in memory?

No dimensionality limits but I wouldn't use this for truly high-dimensional problems like text analysis and inverted indexes (1000s of dimensions). It might work but I wasn't really thinking about those kinds of use cases when building this.