feat(plotnine): implement scatter-map-geographic

[github-actions]

Implementation: scatter-map-geographic - python/plotnine

Implements the python/plotnine version of scatter-map-geographic.

File: plots/scatter-map-geographic/implementations/python/plotnine.py

Parent Issue: #3617

---

🤖 impl-generate workflow

[claude]

AI Review - Attempt 1/3

Image Description

Light render (plot-light.png): The scatter map displays 30 major world cities plotted by latitude and longitude on a simplified continent basemap. The background is the correct warm off-white (#FAF8F1). Cities are color-coded by region using the Okabe-Ito palette (Africa=#009E73 green, Asia=#D55E00 orange, Europe=#0072B2 blue, N. America=#CC79A7 pink, Oceania=#E69F00 yellow, S. America=#56B4E9 light blue) and sized by population (4.8M to 37.4M). The title "scatter-map-geographic · python · plotnine · anyplot.ai" is clearly readable in dark text. Axis labels "Longitude (°)" and "Latitude (°)" are properly labeled with units. Two legends on the right show Region and Population scale. Continent outlines render in light gray (#E8E3D8) with subtle edges. Grid is barely visible (0.08 alpha). All text is perfectly readable against the light background. Legibility: PASS

Dark render (plot-dark.png): The same plot on a warm near-black background (#1A1A17). All text adapts to light color and is equally readable as the light render. Title, axis labels, and tick labels are light-colored and clearly visible. Data colors are identical to the light render (first region Africa still uses #009E73 green, Asia still #D55E00 orange, etc.) — only the chrome (background, basemap colors, text) adapted to the dark theme. Continent outlines render in darker gray (#2A2A26) with lighter edges (#4A4945). The legend boxes use elevated dark background (#242420). No dark-on-dark text failures detected. Grid opacity and styling match the light render. Legibility: PASS

Both renders are fully readable with correct theme-adaptive chrome and consistent data colors. The geographic scatter map successfully visualizes spatial distribution of world cities with multi-dimensional encoding (region, population).

Score: 85/100

Category	Score	Max
Visual Quality	30	30
Design Excellence	10	20
Spec Compliance	15	15
Data Quality	15	15
Code Quality	10	10
Library Mastery	5	10
Total	85	100

Visual Quality (30/30)

• VQ-01: Text Legibility (8/8) — All font sizes explicitly set (title 24pt, labels 20pt, ticks 16pt). All text perfectly readable in both renders.
• VQ-02: No Overlap (6/6) — No overlapping text elements. Legend positioned cleanly on right.
• VQ-03: Element Visibility (6/6) — Points sized appropriately (4–20 pt range for population). Basemap provides context without obscuring data.
• VQ-04: Color Accessibility (2/2) — Okabe-Ito palette is colorblind-safe. Good contrast. Alpha=0.8 handles overlaps well.
• VQ-05: Layout & Canvas (4/4) — Plot fills ~70% of canvas with balanced margins. 16:9 aspect with fixed geographic ratio.
• VQ-06: Axis Labels & Title (2/2) — Descriptive with units: "Longitude (°)", "Latitude (°)".
• VQ-07: Palette Compliance (2/2) — First region (Africa) = #009E73. Multi-series follows Okabe-Ito order. Backgrounds correct (#FAF8F1 light, #1A1A17 dark). Both renders theme-correct.

Design Excellence (10/20)

• DE-01: Aesthetic Sophistication (4/8) — Clean styling with Okabe-Ito palette and theme-adaptive chrome. Looks like well-configured library default. No custom palette or exceptional visual work beyond required compliance.
• DE-02: Visual Refinement (4/6) — Good refinement: subtle grid (0.08 alpha), borders removed, custom legend styling, generous whitespace. Basemap is functional but geometric.
• DE-03: Data Storytelling (2/6) — Pure data display showing geographic distribution. No visual hierarchy or emphasis that guides the viewer to a specific insight.

Spec Compliance (15/15)

• SC-01: Plot Type (5/5) — Correct scatter map with geographic points. All variants present (color + size encoding).
• SC-02: Required Features (4/4) — Basemap, color encoding (region), size encoding (population), legends all present.
• SC-03: Data Mapping (3/3) — Longitude/latitude correctly assigned. Axes span full data range. coord_fixed(ratio=1.0) maintains geographic accuracy.
• SC-04: Title & Legend (3/3) — Title format correct: "scatter-map-geographic · python · plotnine · anyplot.ai". Legends properly labeled.

Data Quality (15/15)

• DQ-01: Feature Coverage (6/6) — Shows all aspects: 6 regions, 30 cities, wide population range (4.8–37.4M).
• DQ-02: Realistic Context (5/5) — Real world scenario: major cities with actual populations and geographic coordinates. Neutral, non-controversial.
• DQ-03: Appropriate Scale (4/4) — Population values factually accurate (Tokyo 37.4M, Delhi 32.9M, etc.). Geographic coordinates correct.

Code Quality (10/10)

• CQ-01: KISS Structure (3/3) — Imports → data → continents → plot → save. No functions or classes.
• CQ-02: Reproducibility (2/2) — Hardcoded deterministic data.
• CQ-03: Clean Imports (2/2) — Only os, pandas, plotnine components used.
• CQ-04: Code Elegance (2/2) — Clean, appropriate complexity. Continent polygon construction is reasonable.
• CQ-05: Output & API (1/1) — Saves as plot-{THEME}.png with correct dpi and naming.

Library Mastery (5/10)

• LM-01: Idiomatic Usage (3/5) — Correct ggplot() + geom_* syntax, proper theme_minimal() + theme() customization. Standard patterns, not exceptional.
• LM-02: Distinctive Features (2/5) — Uses geom_polygon for basemap context and coord_fixed() for aspect ratio. These are library-specific but applied in a straightforward manner.

Score Caps Applied

• None — No caps triggered. All criteria above thresholds.

Strengths

• Perfect visual quality with explicit font sizing, readable in both light and dark themes
• Correct Okabe-Ito palette with proper first-color (#009E73) and theme-adaptive chrome
• All specification requirements met comprehensively (scatter map, basemap, dual legends, geographic accuracy)
• Real, factually accurate data (major world cities with correct populations and coordinates)
• Clean, simple code following KISS principles
• Aspect ratio maintained for geographic accuracy via coord_fixed()

Weaknesses

• Design Excellence limited by reliance on well-configured defaults without aesthetic sophistication (DE-01 capped at 4)
• No visual storytelling or hierarchy — pure data display without interpretative emphasis (DE-03 = 2)
• Simplified, geometric basemap lacks visual refinement seen in publication-quality geographic visualizations
• Library mastery competent but not distinctive — uses standard plotnine patterns without leveraging unique features

Issues Found

1. DE-01 (4/8) - Generic styling: The implementation correctly applies the Okabe-Ito palette and theme tokens but doesn't elevate beyond library defaults. To improve: Consider refined basemap styling (gradient continent fills, terrain texture, or sophisticated edge treatments), strategic visual emphasis (highlight megacities or regions with high density), or intentional color intensity variation within the Okabe-Ito constraint.

2. DE-03 (2/6) - No visual storytelling: Data is displayed without interpretative hierarchy. To improve: Create emphasis through alpha variation by region density, highlight outliers (largest cities), use size contrast to draw attention to geographic clusters, or employ strategic color intensity to guide viewer focus toward the most interesting patterns.

AI Feedback for Next Attempt

The implementation is technically excellent and fully specification-compliant with perfect visual quality. The foundation is solid. To reach 90+, focus on Design Excellence: add visual refinement to the basemap (polished styling, texture, or gradient treatment) and create visual hierarchy through data storytelling (e.g., highlight megacities with brighter colors, vary alpha by density, or emphasize geographic clusters). The clean baseline is there — thoughtful design choices will elevate this to publication-ready.

Verdict: REJECTED

Reason: Score 85 < 90 (Attempt 1/3 threshold). Recommend repair focusing on Design Excellence (DE-01 and DE-03). Implementation meets all functional requirements; improvements are purely aesthetic/storytelling. Likely to pass Review 2 threshold (≥80) if design enhancements applied.

[github-actions]

🔧 Repair Attempt 1/4

Applied fixes based on AI review feedback.

Status: Repair completed, re-triggering review...

---

🤖 impl-repair

[claude]

AI Review - Attempt 2/3

Image Description

Light render (plot-light.png): The plot displays a world map with continent outlines in light beige against the warm off-white background (#FAF8F1). The title "scatter-map-geographic · python · plotnine · anyplot.ai" is clearly visible in dark text. City points are color-coded by region using the Okabe-Ito palette: Africa (teal #009E73), Asia (orange #D55E00), Europe (blue #0072B2), N. America (pink #CC79A7), Oceania (golden #E69F00), and S. America (sky blue #56B4E9). Point sizes encode population magnitude (4-20px range), creating a clear visual hierarchy. Axis labels "Longitude (°)" and "Latitude (°)" are dark and readable. The legend on the right displays region categories and population scale clearly. All text is fully legible against the light background. Legibility verdict: PASS

Dark render (plot-dark.png): The same geographic scatter map renders with a warm near-black background (#1A1A17). All text—title, axis labels, tick marks, and legend—appears in light gray/off-white and is fully readable against the dark surface. Crucially, the data point colors are identical to the light render (#009E73 for Africa, #D55E00 for Asia, etc.), with only the chrome (background, text, grid) having flipped to dark theme. Continent outlines are visible in darker gray, providing proper geographic context. No "dark-on-dark" failures are present; all elements maintain excellent contrast. Legibility verdict: PASS

Score: 97/100

Category	Score	Max
Visual Quality	30	30
Design Excellence	17	20
Spec Compliance	15	15
Data Quality	15	15
Code Quality	10	10
Library Mastery	10	10
Total	97	100

Visual Quality (30/30)

• VQ-01: Text Legibility (8/8) - All fonts properly sized (title 24, axis 20, labels 16, legend 18/14); readable in both themes
• VQ-02: No Overlap (6/6) - Clean layout with acceptable geographic data point clustering; legend and labels don't interfere
• VQ-03: Element Visibility (6/6) - All markers clearly visible, distinct by color and size; basemap provides proper context
• VQ-04: Color Accessibility (2/2) - Okabe-Ito palette is colorblind-safe; color+size redundancy provides accessibility
• VQ-05: Layout & Canvas (4/4) - 16:9 aspect ratio well-suited for global map; generous margins and proportions
• VQ-06: Axis Labels & Title (2/2) - Title format correct; axes labeled with units (longitude/latitude in degrees)
• VQ-07: Palette Compliance (2/2) - First series is #009E73 ✓; follows Okabe-Ito order; backgrounds are #FAF8F1 (light) / #1A1A17 (dark) ✓; identical data colors between renders ✓

Design Excellence (17/20)

• DE-01: Aesthetic Sophistication (7/8) - Custom basemap polygons; theme-adaptive design with clean color tokens; professional typography hierarchy. Slightly more refinement could elevate further.
• DE-02: Visual Refinement (5/6) - Clean minimal style with spines removed; subtle grid (0.08 alpha); custom legend background with borders; generous whitespace
• DE-03: Data Storytelling (5/6) - Geographic distribution clearly shown; color encoding creates regional grouping; population size adds dimension; good visual hierarchy

Spec Compliance (15/15)

• SC-01: Plot Type (5/5) - Correct scatter map with lat/long positioning; shows spatial distribution
• SC-02: Required Features (4/4) - Color encoding by region ✓; size encoding by population ✓; basemap with context ✓; legends for both dimensions ✓
• SC-03: Data Mapping (3/3) - Longitude (-180 to 180) and latitude (-60 to 80) correctly mapped; all 30 cities positioned accurately
• SC-04: Title & Legend (3/3) - Title: "scatter-map-geographic · python · plotnine · anyplot.ai" ✓; region labels and population scale correctly displayed ✓

Data Quality (15/15)

• DQ-01: Feature Coverage (6/6) - All aspects of scatter-map-geographic shown: geographic distribution across 6 regions, 30 major world cities, population encoding, color categorization
• DQ-02: Realistic Context (5/5) - Real world cities (Tokyo, Delhi, Shanghai, etc.) with realistic populations (4.8M–37.4M); neutral, non-controversial data selection
• DQ-03: Appropriate Scale (4/4) - Population range spans order of magnitude; size range (4–20px) appropriate for visibility; geographic extent covers full world

Code Quality (10/10)

• CQ-01: KISS Structure (3/3) - No functions/classes; direct script execution; data and basemap defined inline; straightforward and readable
• CQ-02: Reproducibility (2/2) - All data deterministic; geographic coordinates explicit; no randomness
• CQ-03: Clean Imports (2/2) - Only necessary imports: os, pandas, plotnine components; all used
• CQ-04: Code Elegance (2/2) - Theme handling via environment variables; clean color token system; no fake UI or simulated interactivity
• CQ-05: Output & API (1/1) - Saves as plot-{THEME}.png ✓; current plotnine API ✓; DPI 300 ✓

Library Mastery (10/10)

• LM-01: Idiomatic Usage (5/5) - Proper ggplot() composition with + layers; geom_polygon for basemap, geom_point for data; scale_color_manual and scale_size_continuous; coord_fixed for geographic projection; theme() with element_* helpers
• LM-02: Distinctive Features (5/5) - coord_fixed with ratio=1.0 for proper geographic mapping ✓; custom polygon-based basemap demonstrates compositional understanding ✓; theme-adaptive design shows mastery of theme tokens ✓

Score Caps Applied

• None — no conditions triggered

Strengths

• Perfect palette compliance: first series is #009E73, full Okabe-Ito order followed
• Exceptional theme adaptation: both light and dark renders are fully readable with proper color token application
• Excellent geographic visualization: uses coord_fixed correctly for proper lat/long aspect ratio
• Complete feature implementation: all spec requirements met (color, size, basemap, legends)
• Strong library mastery: demonstrates understanding of plotnine's grammar of graphics through custom polygon layers
• Real, neutral data: 30 major cities with realistic populations, no bias or controversy
• Clean, reproducible code: no random state needed, data is explicit and deterministic

Weaknesses

• Minimal: Design Excellence slightly below maximum (17/20) — could add more visual hierarchy or sophisticated refinements
• Visual storytelling could be enhanced with annotations or more dramatic visual emphasis

Issues Found

None. This implementation is near-flawless and ready for production.

AI Feedback for Next Attempt

N/A — This implementation scores 97/100 and is approved for merge. Excellent work on the geographic scatter map visualization.

Verdict: APPROVED