Module 2 Hands-On Lab: Visual Storytelling, Statistical Distributions & Heatmaps
Transform raw transactional data into publication-ready statistical visualizations using Matplotlib's Object-Oriented interface and Seaborn.
Download Starter Notebook: 02_data_visualization_and_heatmaps_starter.ipynb
Scenario & Business Context
Marketing and product leadership need a visual readout of customer order distributions, category performance, and multi-channel campaign returns. You will design clean, high-resolution figures that avoid misleading visual encodings.
What You Will Build & Demonstrate
| Task | Deliverable | Points |
|---|---|---|
| T13 · Build a shared-axis trend dashboard | fig_trend, two vertically stacked OO subplots, shared x-axis, 300 DPI export | 5 |
| T14 · Show counts, distributions & group shape | fig_distribution, 2x2 panel (KDE histogram, regional countplot, boxplot, strip/violin) | 4 |
| T15 · Distinguish total revenue from Seaborn means | category_summary, fig_category, horizontal bar charts with long labels | 3 |
| T16 · Encode four campaign dimensions | fig_campaign, marketing spend vs revenue, log-scale x, profit margin color gradient | 4 |
| T17 · Build & read executive correlation heatmaps | order_features, corr, masked upper-triangle (np.triu), diverging palette | 4 |
| T18 · Use pairwise, marginal & faceted views | pair_grid, joint_grid, facet_grid small multiples across customer tiers | 3 |
Total Points Available: 23 points (Pass: 80% / 19 pts)
Submission Guidelines
- Download the starter notebook
.ipynb. - Complete all visualization tasks and written chart interpretations.
- Verify all exported PNG figures regenerate in your local workspace.
- Save, restart kernel, execute from top to bottom, and upload below.