Skip to content

Aggregation

The aggregation tab collapses your (possibly irregular) observations into regular time periods. This produces a clean panel with one price (and optionally one quantity) per product per period.

Frequency

Choose how to group observations in time:

Value Meaning
1d Daily
1w Weekly
1mo Monthly
1q Quarterly
1y Yearly

For example, with 1mo, all observations within the same calendar month are combined into a single monthly value per product.

Aggregation type

Select how multiple prices within a period are combined:

Type Description
arithmetic Arithmetic mean of prices.
geometric Geometric mean of prices.
harmonic Harmonic mean of prices.

Weighted variants

If you imported a quantity column, weighted variants become meaningful:

Type Description
weighted_arithmetic Quantity-weighted arithmetic mean.
weighted_geometric Quantity-weighted geometric mean.
weighted_harmonic Quantity-weighted harmonic mean.

Requires quantity data

Weighted aggregation is only available when quantity was mapped at import. The app blocks the run and notifies you if quantities are missing.

Result

After running, the preview shows the aggregated panel with columns:

  • period — the grouped time period (Date),
  • product_id — the product,
  • aggregated_price — the combined price,
  • aggregated_quantity — the combined quantity (only for weighted types).

The info line reports the number of periods, rows, and products.

Continue to Panel Balancing.