Skip to main content

Roll-Up Operations in Analytics

Control how your calculations roll up values to an aggregated result

Holly Conrad Smith avatar
Written by Holly Conrad Smith
Updated over 2 months ago

Analytics always performs the formula on each geometry individually, then operates on the resulting values to generate the final result.

How roll-up Operations Work

The way calculations roll up in your project varies based on the type of components used in the calculation.

Calculations which only use the non-property component types (measure, subtotal, site area) will be calculated once.

When using feature properties as components of your formula, you also have the option of choosing how to determine the method of calculation of the final value of the measure, which is aggregated across all of the affected features.

Calculations which utilise a Property-driven component will always be calculated for each geometry that matches, then rolled up with an operation.

Calculations which are a mix of Property-driven and other component types will be calculated per feature, with the other component types applied in full for each geometry looped through.

When calculating formulas on a feature, geometries that do not have a property value set are treated as Null (not 0).

Operations

After creating a formula in your calculation, you can select the roll-up Operation.

The different roll-up operation options return different results.

  • Sum: Adds all feature results

  • Average: Averages the feature results

  • Minimum: Returns the smallest feature result

  • Maximum: Returns the largest feature result

  • Count: Counts the features

  • Constant: Does not operate on features. Returns the value typed into the formula bar.

Example:

In this example, 2 geometries are passed into analytics.

The formula calculates hard cost. Hard cost is GBA * Hard Cost

The result of the measure depends on the selected Operation:

  • Sum: 39,587,950

  • Average: 19,793,975

  • Minimum: 10,620,232

  • Maximum: 28,967,718

  • Count: 2

Did this answer your question?