Learn what share of demand each size accounts for within a (style × attributes) group, then use it to disaggregate a style-level forecast into size-level forecasts. The library reads from a dataset you already have loaded — it does not require a new upload.
Libraries you've already trained. Click Apply to disaggregate a forecast vector with one.
| Name | Size column | Attribute columns | SKUs | Groups | Obs. | Created | |
|---|---|---|---|---|---|---|---|
| Loading… | |||||||
Picks a dataset and learns the size shares per attribute group. For example, with
sizeColumn = "Size" and attributeCols = ["Category", "Fit"],
every distinct (Category, Fit) combination gets its own size curve.
Splits a style-level forecast into per-size forecasts using a library's coefficients.