The data insights modules contain insights about data in a given data collection. These modules are optional.

To create the data insights modules:

  1. Create a list called Data Collection.
    This list includes the names of the data collections you create in Forecaster.
  2. Create a module with four components, as shown below. This module should be called Time and item-level insights. The components are:
      • Data collection list as a page selector.
      • Time dimension as a row.
        • The dimension should represent monthly, weekly, or daily data.
        • The frequency of the data should correspond to that of the historical data.
      • Item ID list as a row (the same list used in the previous modules).
      • New line items as columns. Create the line items below and format their type in Blueprint mode as detailed below.
        • Is_outlier (Boolean)
        • Is_changepoint (Boolean)
        • Seasonality_weekly (Number)
        • Seasonality_bi_weekly (Number)
        • Seasonality_monthly (Number)
        • Seasonality_quarterly (Number)
        • Seasonality_yearly (Number)
        • Trend_quarterly (Number)
        • Trend_yearly (Number)

You can't use line items that are configured as summary items.

  1. Create a module with three components, as shown below. This module should be called Item-level insights. This module is identical to the previous one, except that it doesn't contain a time dimension.
      • Data collection list as a page selector
      • Item ID list as a row (the same list used in the previous modules)
      • New line items as columns. Create the line items below and format their type in Blueprint mode as detailed below.
        • New_item (Number)
        • Obsolete_item (Number)
        • Stats_count (Number)
        • Stats_mean (Number)
        • Stats_standard_dev (Number)
        • Stats_min (Number)
        • Stats_25th_percentile (Number)
        • Stats_median (Number)
        • Stats_75th_percentile (Number)
        • Stats_maximum (Number)
        • Stats_percentage_total (Number)
        • Stats_skew (Number)
        • [Related_data_line_item_name] + ‘_correlation’ (Number) - only create this line item type if related data is available
        • Sparsity_percentage (Number)
        • Variability_type (Text)

Example of an item-level, insights module setup.

The setup for a new Item-level Insights module with dimensions.

Example of a time and item-level, Insights module setup.

The setup for a time and item-level Insights module with dimensions.