feat(plotnine): implement scatter-marginal

[github-actions]

Implementation: scatter-marginal - python/plotnine

Implements the python/plotnine version of scatter-marginal.

File: plots/scatter-marginal/implementations/python/plotnine.py

Parent Issue: #2005

---

🤖 impl-generate workflow

[claude]

AI Review - Attempt 1/3

Image Description

Light render (plot-light.png): The plot displays a warm off-white (#FAF8F1) background with a composite layout showing scatter plot (lower-left), top histogram (upper-left), and right histogram (upper-right) with a spacer in the corner. Title "scatter-marginal · plotnine · anyplot.ai" in dark text (24pt) is clearly readable at the top. The main scatter plot shows 200 points in brand green (#009E73) with 0.6 alpha transparency, displaying a positive correlation between Study Hours per Week (X-axis, 20pt label) and Exam Score (%) (Y-axis, 20pt label). Tick labels at 16pt are clearly visible in dark text. Marginal histograms use vermillion (#D55E00) with 0.7 alpha and 15 bins each. Subtle grid lines (10% opacity) are visible but non-dominant. All text is readable against the light background with dark color (#1A1A17), confirming proper INK token application. Legibility verdict: PASS

Dark render (plot-dark.png): The plot displays a warm near-black (#1A1A17) background with identical layout. Title in light text (#F0EFE8) is clearly readable. Axis labels and tick labels in soft light text (#B8B7B0) are all clearly visible with high contrast against the dark background. The scatter points remain in brand green (#009E73)—identical color to light render, confirming data colors persist across themes. Marginal histograms in vermillion (#D55E00) show the same distributions. Grid lines are subtle and visible. No dark-on-dark contrast issues detected; all text is readable. Brand green (#009E73) is clearly visible and accessible on the dark surface. Legibility verdict: PASS

Score: 89/100

Category	Score	Max
Visual Quality	30	30
Design Excellence	13	20
Spec Compliance	15	15
Data Quality	15	15
Code Quality	10	10
Library Mastery	6	10
Total	89	100

Visual Quality (30/30)

• VQ-01: Text Legibility (8/8) - Title 24pt, axis labels 20pt, tick labels 16pt—all readable in both themes with proper color tokens
• VQ-02: No Overlap (6/6) - Top histogram, spacer, main scatter, and right histogram compose cleanly without overlap
• VQ-03: Element Visibility (6/6) - 200 points with 0.6 alpha reveal density; 15-bin histograms clearly visible with distinct colors
• VQ-04: Color Accessibility (2/2) - Okabe-Ito palette provides colorblind-safe contrast; no red-green sole signal
• VQ-05: Layout & Canvas (4/4) - 4800×2700 landscape format with balanced proportions; nothing cut off
• VQ-06: Axis Labels & Title (2/2) - Title matches spec format; X and Y labels descriptive with units
• VQ-07: Palette Compliance (2/2) - Scatter #009E73 (position 1), marginals #D55E00 (position 2), backgrounds correct (#FAF8F1 light / #1A1A17 dark), data colors identical across themes

Design Excellence (13/20)

• DE-01: Aesthetic Sophistication (6/8) - Okabe-Ito palette and layout show intentionality; relies on theme_minimal() defaults without custom refinement like spine removal
• DE-02: Visual Refinement (3/6) - Grid is subtle, whitespace generous; top/right spines remain visible (could be removed); borders functional but not refined
• DE-03: Data Storytelling (4/6) - Layout clearly shows bivariate relationship and marginals with good hierarchy; focal point on scatter is clear

Spec Compliance (15/15)

• SC-01: Plot Type (5/5) - Correctly implements scatter plot with marginal distributions (histograms)
• SC-02: Required Features (4/4) - Main scatter + top marginal (x) + right marginal (y); axes align; moderate transparency (0.6/0.7); subtle marginal colors
• SC-03: Data Mapping (3/3) - X: Study Hours (0–50); Y: Exam Score (30–105); axes show all 200 points without clipping
• SC-04: Title & Legend (3/3) - Title format correct; single-series plot requires no legend

Data Quality (15/15)

• DQ-01: Feature Coverage (6/6) - Displays scatter plot, x distribution, y distribution; 200 points in recommended range
• DQ-02: Realistic Context (5/5) - Study hours and exam scores real-world plausible; positive correlation realistic; data neutral
• DQ-03: Appropriate Scale (4/4) - Study hours 0–50 sensible; exam scores 30–105 realistic; 15 bins appropriate for 200 points

Code Quality (10/10)

• CQ-01: KISS Structure (3/3) - Straightforward code, no unnecessary functions/classes
• CQ-02: Reproducibility (2/2) - Uses np.random.seed(42) for deterministic data
• CQ-03: Clean Imports (2/2) - All imports used; no unused dependencies
• CQ-04: Code Elegance (2/2) - No fake UI or simulated interactivity; appropriate complexity
• CQ-05: Output & API (1/1) - Saves as plot-{THEME}.png with correct API

Library Mastery (6/10)

• LM-01: Idiomatic Usage (4/5) - Uses ggplot2 grammar properly; + operator and theme() correct; composition idioms (| /) used well
• LM-02: Distinctive Features (2/5) - Uses composition layout and element_* theming; lacks distinctive plotnine-specific features beyond standard API

Score Caps Applied

• None

Strengths

• Correct Okabe-Ito palette usage (scatter #009E73, marginals #D55E00)
• Proper theme token implementation with correct light/dark colors
• Well-proportioned layout with smooth composition
• Clean readable text in both light and dark renders
• Effective alpha blending (0.6 scatter, 0.7 histograms) reveals density patterns
• Deterministic seeding for reproducibility

Weaknesses

• Design excellence relies on defaults rather than custom refinement
• Top/right spines could be removed for more polished appearance
• No distinctive plotnine features beyond standard composition

Issues Found

1. DE-01 MODERATE: Relies on theme_minimal() defaults without custom enhancements
   • Fix: Remove top/right spines with theme(axis_line=...), customize axis appearance
2. DE-02 MODERATE: Visual refinement could be enhanced
   • Fix: Set axis_line for left/bottom only, refine border appearance around subplots
3. LM-02 LOW: Missing distinctive plotnine features
   • Fix: Could leverage faceting patterns or custom scale transformations for future implementations

AI Feedback for Next Attempt

This is a solid, well-executed implementation that meets all specifications and has excellent theme support and color correctness. To improve design excellence: remove top and right spines, refine the subplot borders, and consider more sophisticated visual customization beyond the default theme. Library mastery can be strengthened by exploring plotnine-specific features like advanced theming or faceting patterns.

Verdict: APPROVED

[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 scatter-marginal visualization with a warm off-white background (#FAF8F1). The title "scatter-marginal · plotnine · anyplot.ai" appears in bold dark text at the top. The L-shaped layout consists of: (1) a top histogram showing the distribution of Study Hours per Week in orange bars (#D55E00), (2) a main scatter plot in the lower-left with 200 teal-green points (#009E73) showing a clear positive correlation between study hours (x-axis: 0-50) and exam scores (y-axis: 30-105%), with appropriate alpha transparency (0.6) revealing density, and (3) a right-side histogram showing the distribution of Exam Scores also in orange bars. Axis labels are bold and readable ("Study Hours per Week" and "Exam Score (%)"), tick labels are visible in a softer gray (#4A4A44), and subtle horizontal grid lines provide visual reference without competing with data. All text is clearly readable against the light background. Legibility verdict: PASS

Dark render (plot-dark.png): The same visualization appears on a warm near-black background (#1A1A17). All text elements have adapted to light colors—the title and axis labels are now white/cream colored (#F0EFE8) and remain perfectly readable against the dark surface. Tick labels are in the appropriate soft gray for dark themes (#B8B7B0). Critically, the scatter points remain the identical teal-green (#009E73) and the histogram bars remain the identical orange (#D55E00) as the light render, confirming that only the chrome elements (background, text, grid) have adapted while the data visualization maintains perfect color consistency. Grid lines are similarly subtle. No text exhibits "dark-on-dark" coloring issues—all chrome is properly adapted. Legibility verdict: PASS

Score: 93/100

Category	Score	Max
Visual Quality	30	30
Design Excellence	16	20
Spec Compliance	15	15
Data Quality	15	15
Code Quality	10	10
Library Mastery	7	10
Total	93	100

Visual Quality (30/30)

• VQ-01: Text Legibility (8/8) - All text properly sized (24pt title, 20pt labels, 16pt ticks), correctly colored for both themes, no readability issues
• VQ-02: No Overlap (6/6) - L-shaped composite layout effectively prevents all overlaps, proper spacing between all elements
• VQ-03: Element Visibility (6/6) - All 200 scatter points visible with density-appropriate alpha (0.6), histogram bars clearly distinguishable
• VQ-04: Color Accessibility (2/2) - Okabe-Ito palette (#009E73 and #D55E00), CVD-safe, adequate contrast, no red-green sole signal
• VQ-05: Layout & Canvas (4/4) - Well-proportioned 16:9 landscape, nothing cropped, generous whitespace, balanced composition
• VQ-06: Axis Labels & Title (2/2) - Title matches spec format, labels descriptive with units ("per Week", "(%)")
• VQ-07: Palette Compliance (2/2) - First series #009E73 (green), second #D55E00 (orange), backgrounds #FAF8F1/#1A1A17, both renders theme-correct

Design Excellence (16/20)

• DE-01: Aesthetic Sophistication (6/8) - Intentional L-shaped layout with custom figure sizing per panel (main_w/main_h vs marg_w/marg_h), clear color hierarchy (orange frames, green focal point), but execution is straightforward without "wow" factor
• DE-02: Visual Refinement (5/6) - Minimal spines (theme_minimal), subtle grid (alpha=0.10), generous margins, histogram edges visible; could be slightly more polished
• DE-03: Data Storytelling (5/6) - Layout naturally guides viewer from marginals to main scatter, correlation visually apparent, could benefit from subtle emphasis

Spec Compliance (15/15)

• SC-01: Plot Type (5/5) - Correct scatter plot with marginal distributions
• SC-02: Required Features (4/4) - Main scatter plot, top histogram (x), right histogram (y), aligned axes
• SC-03: Data Mapping (3/3) - X (study hours 0-50), Y (exam scores 30-105), full data range captured
• SC-04: Title & Legend (3/3) - Proper title format, no legend needed (single series)

Data Quality (15/15)

• DQ-01: Feature Coverage (6/6) - Complete scatter-marginal with joint and marginal distributions
• DQ-02: Realistic Context (5/5) - Study hours vs exam scores, realistic ranges, positive correlation is real-world plausible
• DQ-03: Appropriate Scale (4/4) - Sensible axis ranges and data density

Code Quality (10/10)

• CQ-01: KISS Structure (3/3) - Direct code, no unnecessary functions or classes
• CQ-02: Reproducibility (2/2) - np.random.seed(42) ensures deterministic output
• CQ-03: Clean Imports (2/2) - Only necessary imports (os, numpy, pandas, plotnine)
• CQ-04: Code Elegance (2/2) - No fake interactivity, appropriate complexity, plotnine composition syntax used idiomatically
• CQ-05: Output & API (1/1) - Saves as plot-{THEME}.png with correct API

Library Mastery (7/10)

• LM-01: Idiomatic Usage (4/5) - Grammar of graphics syntax, proper theme composition with + operator, plot_spacer() for layout, figure_size configuration per panel
• LM-02: Distinctive Features (3/5) - Uses plotnine composition operators (| and /) for L-shaped layout, plot_spacer() for alignment

Score Caps Applied

• None applied - no criteria triggered cap conditions

Strengths

• Perfect theme adaptation: both renders visually correct with zero legibility issues
• Clean, minimal design with intentional L-shaped composite layout
• Proper use of Okabe-Ito palette with excellent color hierarchy
• Well-structured code with proper seed for reproducibility
• Complete spec compliance with all required features
• Excellent axis alignment between scatter plot and marginal distributions

Weaknesses

• Design could show more sophistication or custom refinement beyond standard plotnine defaults
• Code uses straightforward library patterns without pushing plotnine's more distinctive features

AI Feedback for Next Attempt

This is a strong implementation with excellent fundamentals. Both light and dark renders are flawless in terms of readability and theme adaptation. Consider for future iterations: subtle enhancements to visual hierarchy (emphasis on correlation slope), custom annotation of key patterns, or more distinctive styling that goes beyond the minimal defaults. The L-shaped layout is well-executed and the palette choices are perfect.

Verdict: APPROVED