The explainability results module gives insight into which underlying features have the greatest impact on forecast results.

To create the module:

  1. Create a list called Forecast Feature Names.
  2. Create a module called Forecast Explainability Results.
  3. From the New Module dialog:
    1. Use the Item ID list and Forecast Feature Names list as row dimensions.
    2. Use Line Items as a column dimension. 
    3. Use the Forecast Version list as a page dimension (see screenshot 2 below). This is the same list you used in the forecast results module.
    4. Select OK.
  4. Add a line item called Feature Contribution. Format this line item as a number with percentage as the unit. You can't use line items that are configured as summary items.
  5. Add a line item called Applicable Forecast. Format this line item as a Boolean.

Example of an explainability results module:

An example of an explainability results module.