HierarchicalForecast

Do reconciled forecasts satisfy all aggregation constraints in a small synthetic hierarchy?

Construct the summing matrix independently, recompute aggregation residuals, and compare held-out forecast errors separately from coherence.

What you could produce

• A pinned, isolated reproducer with generated inputs and a concise result table
• A separately controlled check report with fixed tolerances and disclosed limitations

Before you use it

• A separately qualified runtime with the package and its reviewed, pinned dependency closure; no dependency installation is supported by the current self-contained Python pilot.

Limits to keep in view

• No project source, package build hook, test, example, or submitted research command has been executed.
• The documented pilot supports self-contained Python 3.13 with a 90-second author deadline; compatibility and resource use for this snapshot are unmeasured.

Source and permission context

Preserve the upstream project name, version, repository link, applicable notices, and contributor attribution when preparing an artifact for reuse.

Catalog listing reviewed. This review covers the description and source links displayed here.

Approved for catalog metadata and links. The displayed entry identifies HierarchicalForecast with the concise collection-authored summary “Forecast reconciliation and evaluation for hierarchical and grouped time series.” and points to the public upstream repository https://github.com/Nixtla/hierarchicalforecast. The wording describes function and possible investigation without reproducing upstream source or documentation, claiming execution, or implying endorsement or rights beyond the recorded scopes.

Reviewed 2026-09-14. Copying or adapting source files remains subject to their own terms.

Before copying source material

• Bundled datasets, examples, submodules, vendored code, and dependency licenses have not been audited; the observed top-level license does not clear all of them.
• README and documentation rights were not independently resolved from the main code license; this catalog only links and describes.