Forecasting Intelligence  Lab

Datasets

Two independent retail environments

M5 Walmart is the primary dataset; Corporación Favorita provides replication. Findings across them are reported as replication and divergence, not generalization.

Datasets

Two independent retail demand environments.

M5 is the primary dataset; Corporación Favorita is used for replication and divergence across retail environments.

Primary dataset

M5 — Walmart Retail Sales

Product × Store × Day
  • Daily retail demand at product–store level
  • Calendar variables and event days
  • Item prices and price changes
  • SNAP (supplemental assistance) indicators
  • Multiple demand patterns across series

Primary dataset for the full experiment.

Validation / Replication dataset

Corporación Favorita Grocery Sales

Product × Store × Day
  • Daily grocery sales by store and product
  • Promotion indicators
  • Holiday and event information
  • Store and product characteristics
  • A second, structurally different retail environment

Independent retail environment used for replication.

M5 and Favorita differ in geography, retailer, product structure, promotions, scale and demand patterns. They are not controlled replications — findings are reported as replication and divergence, not as proof of generalization.

Sampling strategy

A representative, stratified sample.

The study does not process every series. A representative sample — approximately 100–300 product×store series per dataset — is selected, stratified by demand segment (stable, trending, seasonal, intermittent). The LLM experiment runs on a smaller, compute-bounded subsample.

Both datasets are structured as product × store × day, enabling the same pipeline to run on each. Raw data files are never committed to the repository; download instructions are documented in data/README.md.