dashboard-beautifylisted

# Data Atelier Production-grade dashboards and infographics with Dash and Plotly Python. **Philosophy**: form follows function, colour communicates, every design decision earns its place. A dashboard is a curated argument made from data — not a chart dump. **Default target**: production delivery. Optimise for correctness, maintainability, and accessibility from the start — not as an afterthought. --- ## Before writing any code — establish the brief 1. **What is the primary question this dashboard answers?** The answer must be visible in < 3 seconds. 2. **Who is the audience?** Executive / analyst / operational — determines density, annotation level, interactivity depth. 3. **Single infographic or multi-panel dashboard?** Different layout and scope strategies apply. 4. **What is the data shape?** Time series / categories / relationships / distributions / compositions. 5. **What is the data source and refresh cadence?** Determines caching strategy and performance approach. If any of these are unanswered, ask before writing code. --- ## Step 1 — Data quality gate Run this before any design work. A beautiful dashboard built on bad data is worse than no dashboard. Read `references/data-quality.md` for full checks. Minimum required: - **Missingness**: identify null rates per column — flag anything > 5% to the user before proceeding - **Duplicates**: check for duplicate rows on the natural key; deduplicate or explain why not - **Outliers**: surface extreme values — confi