Urban Transit Lab
Home Forecast Comparison Research Roadmap Weather GitHub
Resources

Research Library

Forecast guides, source inventories, project cards, and comparable analysis charts. Each artifact remains labeled by source so its provenance is visible.

Reference Artifacts · August 2026

Forecast Source Guides & Inventories

Working artifacts document the event, holiday, ISO-week, and weather context used for demand planning. ChatGPT and Claude reference sets are retained separately so their methodological provenance remains visible for comparison.

ChatGPT · Field Edition

ChatGPT Forecast

Branded operating guide with event and holiday separation, ISO-week correlation, historical weather patterns, projected weather-risk caveats, and source methodology for demand forecasting.

Word guide → Excel inventory →
Claude · Reference Edition

Claude Forecast

Companion source guide and historical event workbook retained as supplied for methodological comparison and reference. Includes event records, ISO-week series, projected weeks, and joining logic.

Word guide → Excel inventory →
ChatGPT Forecast project card summarizing canonical events, holidays, weather-risk days, and ISO-week methodology
ChatGPT Forecast card — Event-driven demand methodology
Claude Forecast project card summarizing event records, ISO-week series, projected weeks, and joining methodology
Claude Forecast card — Historical event series & projections
ChatGPT Analysis chart showing confirmed or logged events, projected additional events, and same-week baselines for ISO weeks 32–53 of 2026
ChatGPT Analysis — Workbook-derived event projection (weeks 32–53)
Claude Analysis chart showing confirmed or logged events, projected additional events, and same-week baselines for ISO weeks 32–53 of 2026
Claude Analysis — Workbook-derived event projection (weeks 32–53)
Events Database

Integrated Events Viewer

135 events from ChatGPT Forecast (90) and Claude Forecast (45) models (2024–2026). Search, filter by date range or source, and view ISO week assignments.

to
ID Event Name Category Area Impact Src ISO YR-WK Dates

Loading... events · All sources

About This Library

Methodology & Data Sources

These guides and charts represent two independent forecasting approaches applied to the same Columbus GBFS fleet data. Both use historical event catalogs, ISO-week correlations, and weather signals to project demand. The side-by-side comparison enables validation and cross-methodological review.

Data lineage: All forecasts derive from public GBFS feeds (Spin and Veo), publicly available holiday/event calendars, and historical weather records. No proprietary data. Full methodologies documented in the Word guides.