feat(matplotlib): implement dumbbell-basic

[github-actions]

Implementation: dumbbell-basic - python/matplotlib

Implements the python/matplotlib version of dumbbell-basic.

File: plots/dumbbell-basic/implementations/python/matplotlib.py

Parent Issue: #945

---

🤖 impl-generate workflow

[claude]

AI Review - Attempt 1/3

Image Description

Light render (plot-light.png): The plot appears on a warm off-white #FAF8F1 background with 8 horizontal dumbbells, one per department. Categories run along the y-axis (HR at top, Sales at bottom), and the x-axis spans satisfaction scores from 30 to 100. "Before" dots are rendered in Okabe-Ito green (#009E73) and "After" dots in vermillion (#D55E00); each pair is connected by a subtle gray line with alpha=0.55. Dots carry a theme-adaptive white halo (edgecolors=PAGE_BG). The title, x-axis label, y-axis label, tick labels, and legend text are all clearly readable in dark ink against the off-white background. Subtle vertical grid lines aid value reading. The legend sits in the lower right with a warm elevated-BG frame.

Dark render (plot-dark.png): The same chart on a warm near-black #1A1A17 background. Data colors are identical — #009E73 green for Before, #D55E00 vermillion for After — only chrome flips. Title, axis labels, category labels on the y-axis, and x-axis tick labels are all rendered in light colors and are clearly readable against the dark surface. No dark-on-dark failures detected. The connecting lines read as medium gray against the dark background. The legend frame uses the elevated dark background (#242420) with a light edge, clearly distinguishable. Grid lines are subtly lighter than the background and do not compete with the data.

Both renders are theme-correct. Legibility passes in both light and dark modes.

Score: 85/100

Category	Score	Max
Visual Quality	28	30
Design Excellence	12	20
Spec Compliance	15	15
Data Quality	15	15
Code Quality	10	10
Library Mastery	5	10
Total	85	100

Visual Quality (28/30)

• VQ-01: Text Legibility (7/8) — Font sizes explicitly set (title=24, labels=20, ticks=16), all text readable in both themes. Minor: y-tick fontsize relies on global tick_params rather than per-axis sizing.
• VQ-02: No Overlap (6/6) — All elements well-spaced, no collisions with 8 categories
• VQ-03: Element Visibility (6/6) — Dot size s=320 is optimal for 8 data points; connecting lines at linewidth=2.5 with alpha=0.55 are visible but appropriately subtle
• VQ-04: Color Accessibility (2/2) — Okabe-Ito palette is CVD-safe; green vs. vermillion distinguishable by luminance difference, not hue alone
• VQ-05: Layout & Canvas (4/4) — 16:9 canvas well-utilized, tight_layout applied, legend near data, nothing cut off
• VQ-06: Axis Labels & Title (1/2) — "Satisfaction Score" is descriptive but lacks units/scale (e.g., "(0–100)")
• VQ-07: Palette Compliance (2/2) — First series is #009E73, second is #D55E00; backgrounds correct (#FAF8F1 / #1A1A17); full theme-adaptive chrome

Design Excellence (12/20)

• DE-01: Aesthetic Sophistication (5/8) — Uses correct Okabe-Ito palette, theme-adaptive dot halos (edgecolors=PAGE_BG), data sorting for narrative effect, and clean spine removal. Above a pure default but not yet at publication-ready level.
• DE-02: Visual Refinement (4/6) — Top/right spines removed, x-axis grid is subtle (alpha~0.18), connecting lines muted with alpha=0.55, legend frame styled. Clear visible refinement over library defaults.
• DE-03: Data Storytelling (3/6) — Sorting by improvement (HR at top with largest gap, Sales at bottom) reveals the data story. However, there is no visual emphasis on the top/bottom performer — the viewer must identify the insight themselves.

Spec Compliance (15/15)

• SC-01: Plot Type (5/5) — Correct dumbbell/connected dot chart type
• SC-02: Required Features (4/4) — Horizontal orientation, distinct Before/After colors, sort by difference, subtle connecting line
• SC-03: Data Mapping (3/3) — Categories on y-axis, satisfaction values on x-axis, all data visible
• SC-04: Title & Legend (3/3) — Title: "dumbbell-basic · matplotlib · anyplot.ai"; legend labels "Before" / "After"

Data Quality (15/15)

• DQ-01: Feature Coverage (6/6) — Shows 8 departments with varying Before/After gaps, both small and large improvements, full range of dumbbell distances
• DQ-02: Realistic Context (5/5) — Employee satisfaction scores before/after workplace policy changes — realistic, neutral business scenario
• DQ-03: Appropriate Scale (4/4) — Values 40–88 on a 30–100 axis; plausible for a 0–100 satisfaction scale

Code Quality (10/10)

• CQ-01: KISS Structure (3/3) — No functions or classes; clean Imports → Data → Sort → Plot → Save flow
• CQ-02: Reproducibility (2/2) — Fully deterministic hardcoded data
• CQ-03: Clean Imports (2/2) — Only os, matplotlib.pyplot, numpy — all used
• CQ-04: Code Elegance (2/2) — Clean Pythonic code; zip with strict=True; no fake UI elements
• CQ-05: Output & API (1/1) — Saves as plot-{THEME}.png with facecolor; current matplotlib API

Library Mastery (5/10)

• LM-01: Idiomatic Usage (3/5) — Correct Axes API throughout; legend frame manipulation; tick_params; set_axisbelow for proper grid ordering
• LM-02: Distinctive Features (2/5) — Theme-adaptive edgecolors=PAGE_BG for dot halos is a nice matplotlib-specific touch; zorder control is used intentionally. Could go further with custom legend handles or PathEffects.

Score Caps Applied

• None — all cap conditions are clear (VQ-02>0, VQ-03>0, SC-01>0, DQ-02>0, DE-01=5>2, CQ-04>0)

Strengths

• Full theme-adaptation: backgrounds, text, connecting lines, legend frame, and dot halos all flip correctly between light and dark
• Correct Okabe-Ito palette with Before=#009E73 as first series
• Data sorting by improvement size creates a natural visual narrative
• Connecting lines are intentionally muted (alpha=0.55) so they don't overpower the dots, per spec recommendation
• Deterministic, realistic, neutral example data (employee satisfaction across 8 departments)
• Perfect code structure and all spec requirements met

Weaknesses

• X-axis label "Satisfaction Score" lacks a scale indicator — adding "(0–100)" would satisfy the units rubric
• No visual emphasis on notable data points — the highest-improvement department (HR, +28 pts) and others receive identical visual treatment; color intensity, size variation, or a subtle callout would elevate data storytelling
• Library Mastery could be stronger — auto-generated legend from scatter labels rather than custom handles misses a matplotlib-specific opportunity for more control
• DE-03 is capped by missing visual hierarchy beyond sorting: consider a subtle annotation or dot-size variation for emphasis

Issues Found

1. VQ-06 LOW: X-axis label missing scale context
   • Fix: Change to or
2. DE-03 MEDIUM: Sorting reveals the story but provides no visual emphasis
   • Fix: Consider coloring the connecting line for the top 1-2 most-improved rows in a third Okabe-Ito accent, or varying line thickness by improvement size, or adding a small value difference label on the widest gap
3. LM-02 LOW: Auto-legend from scatter labels, no custom handles
   • Fix: Use pattern or for custom legend entries to demonstrate more matplotlib control

AI Feedback for Next Attempt

Improve on three fronts: (1) Add "(0–100)" to the x-axis label for units compliance. (2) Elevate data storytelling beyond sorting — highlight the widest gap or most improved department using a thicker/colored connector or a difference label like "+28 pts" on the HR row; this creates a focal point without breaking the clean aesthetic. (3) Use explicit custom legend handles () to show more deliberate matplotlib mastery.

Verdict: REJECTED