Build a GDP Nowcast Release-Impact Workbench

A national-accounts forecasting team.

Update a quarterly GDP nowcast as monthly releases arrive. Attribute changes to new information, historical revisions, and model refits without leaking future observations.

Key: target_quarter + forecast_origin + model_version.

Fields: predictor_period, released_at, vintage_id, transformed_value, forecast, target_vintage.

A missing recent month is unavailable data, not a zero observation.

Build a release-aware monthly predictor panel.

Fit and archive a baseline at each forecast origin.

Recalculate with new data, revised history, and refitted parameters separately.

Export the nowcast change bridge and forecast interval.

A release timeline with nowcast revisions, component attribution, and model-version comparisons.

A scheduled offline nowcast updater with replay support.

A report comparing first-release and later-vintage forecast errors.

Runnable local application, setup commands, and a dependency lockfile.

Versioned input data, data dictionary, acquisition instructions, and rights notes.

Calculation traces, source references, assumption records, and exported review artifacts.

A deterministic baseline and tests for normal, incomplete, contradictory, and failure-recovery cases.

A recorded complete workflow with a reproducible manifest and explicit limitations.

Transformations use only information available at the forecast origin.

The ordered change bridge reconciles exactly, with its order disclosed.

Final revisions are not used as historical training targets unless labeled.

Report performance against a fixed simple baseline on chronological holdouts.

A new release changes the forecast while the model remains fixed. Separate that change from the later parameter refit.

The attached starter files are prototype inputs. Expected starter answers are visible. They are not independent holdouts.

Full target: Acquire eight years of quarterly GDP vintages and monthly predictors. Preserve missing recent months. Start with a bridge regression before a mixed-frequency model.

Complete target contracts and expert review remain pending. Public observations are not historical release vintages.

statsmodels, DuckDB, Plotly

Use an available model appropriate to the task. Record provider, model identifier, version or retrieval date, settings, cost, and latency. Use deterministic code for calculations. Models may extract, retrieve, or explain evidence. They must not invent missing records, override approval rules, or claim professional authority.

Attribution order matters. A nowcast interval is not a guarantee.

No fabricated records, source citations, professional endorsements, or expert-adjudication claims.

No future-data leakage into dated decisions, model fitting, or historical valuation inputs.

No unauthorized payments, account restrictions, regulatory filings, or production changes.

No automated appraisal certification, tax-compliance determination, or unsupported fraud accusation.

No concealed uncertainty or silently invented inputs required for a numerical conclusion.

Correct domain calculations and justified methods with documented assumptions. Weight: 30.

Source lineage, temporal integrity, rights checks, and explicit synthetic-data labels. Weight: 25.

A working application with review, export, error handling, and reproducible state. Weight: 20.

Adversarial cases, uncertainty, missing-data behavior, and sensitivity analysis. Weight: 15.

Evidence-linked conclusions that distinguish facts, assumptions, and unresolved judgments. Weight: 10. Score each dimension from zero to two. Weighted score = sum(weight * score / 2). A failed safety gate blocks acceptance.