OLAP Data Cube Explorer — Slice, Dice, Roll-Up, Drill-Down

Time Granularity

Aggregation Function

Slice

Dice

Hide cells below 0% of max value

Live Aggregates

Aggregate of visible cells (SUM)

—

Visible cells

64 / 64

Granularity

Month (12)

low valuehigh value

X — Product (Electronics/Clothing/Food/Books) Y — Region (North/South/East/West) Z — Time period

How it works

An OLAP data cube stores a measure (here, unit sales) against several dimensions — Product × Region × Time. Analysts explore it with four classic operations, all live in this scene:

Roll-up:    fine grain  → coarse grain  (Month → Quarter),
            aggregate(months in quarter) using SUM/AVG/MAX/MIN
Drill-down: coarse grain → fine grain   (Quarter → Month)
Slice:      fix one dimension to a single value,
            cube[fixed_axis = v] → a 2-D cross-section
Dice:       keep only cells whose measure passes a
            predicate (here: value ≥ threshold × max)

Every cube shown here is aggregated live: roll-up recomputes each quarter's value from its three underlying months with the chosen function — SUM totals them, AVG averages them, MAX/MIN take the extreme month. Slicing dims every cell off the fixed plane to a small marker instead of deleting it, so you can see where the cross-section sits inside the full cube. Dicing does the same for any cell whose value falls under the threshold. The "Aggregate of visible cells" readout re-runs the same function over whatever is currently on screen, so you can watch SUM/AVG/MAX/MIN diverge as you slice and dice.

This is the same mechanism behind real OLAP engines (SQL Server Analysis Services, Druid, ClickHouse's cube-like rollups) that let a data-mining pipeline pre-aggregate billions of rows so a dashboard query answers in milliseconds instead of scanning raw fact tables.

⚠ Couldn't load the 3D engineThree.js failed to load from the CDN. Check your connection and reload.