Skip to main content

AB Controls Tool Icon AB Controls Tool

The AB Controls tool matches one to ten control units (for example, stores, customers, etc.) to each member of a set of previously selected test units, based on criteria such as seasonal patterns and growth trends for a key performance indicator (for example, sales or traffic), along with other user-provided criteria (for example, trade area or demographic characteristics). Possible control units are matched to a user-provided set of treatment units using nearest neighbor methods based on a kd-tree algorithm.

Important

To greatly improve computational speed, the AB Controls tool uses an approximation algorithm to handle the assignment of control units to treatment units. That algorithm performs best when you have a small number of treatment units and a large number of number of control units. But the algorithm isn't able to handle cases where the ratio of treatment units to control units nears 1:1 or where you have many treatment units. In those cases, the tool overrides the option to Allow each control unit to be assigned to only a single treatment unit.

This tool uses the R tool. Go to Options > Download Predictive Tools and sign in to the Alteryx Downloads and Licenses portal to install R and the packages used by the R Tool. See Download and Use Predictive Tools.

Connect Inputs

  • D anchor: (Data) A data stream that contains the unit-level measures to match control units to treatment units.

  • T anchor: (Treatments) A data stream that contains a field that provides the unit identifiers (for example, stores or customers) for a test's treatment units.

Configure the Tool

Use the Base Configuration tab to set the controls for the AB test.

  • The identifier field for the treatment units: The field with the treatment unit identifier for the Treatments input. This field must be some sort of character-based type like String or V-String.

  • The unit identifier field for the selection measures: The field with the unit identifier for the Data input of measures on which to base the match of controls to treatments.

  • Select the numeric measure to match control units to treatment units: The fields in the Data input upon which to base the match of treatments to controls.

  • Number of control units per treatment unit: The number of control units to match to each treatment unit, which can range from 1 to 20.

  • Allow each control unit to be assigned to only a single treatment unit: A checkbox option to make each control assigned to only a single treatment unit, by default this is true.

Use the Graphics Options tab to set the controls for the graphical output.

  • Plot size: Select inches or centimeters for the size of the graph.

  • Graph resolution: Select the resolution of the graph in dots per inch: 1x (96 dpi), 2x (192 dpi), or 3x (288 dpi).

    • Lower resolution creates a smaller file and is best for viewing on a monitor.

    • Higher resolution creates a larger file with better print quality.

  • Base font size (points): Select the size of the font in the graph.

View the Output

  • C (Controls): A Designer data stream with the field Controls, the unit identifier code for each control unit.

  • A (Assignment): A Designer data stream with the field Treatment that gives the unit identifier code for each treatment unit, and the fields Control 1 to Control N (where N is the user-provided number of controls per treatment unit that given the unit identifier code for each control unit assigned to a particular treatment unit, and is provided for diagnostic purposes.