feat(plotnine): implement scatter-regression-lowess

[github-actions]

Implementation: scatter-regression-lowess - python/plotnine

Implements the python/plotnine version of scatter-regression-lowess.

File: plots/scatter-regression-lowess/implementations/python/plotnine.py

Parent Issue: #2855

---

🤖 impl-generate workflow

[claude]

AI Review - Attempt 1/3

Image Description

Light render (plot-light.png): The plot displays a scatter chart with 150 data points representing crop yield vs. temperature on a warm off-white background (#FAF8F1). The scatter points are rendered in brand green (#009E73) with alpha=0.6 transparency, allowing the density distribution to be clearly visible. The LOWESS regression curve is overlaid in a contrasting vermillion/orange (#D55E00) and appears smooth and well-fitted to the data pattern. The title "scatter-regression-lowess · plotnine · anyplot.ai" is clearly visible at the top in dark text. Axis labels are professional: "Temperature (°C)" on the x-axis and "Crop Yield (tons/hectare)" on the y-axis. Tick labels are legible in dark gray. Grid lines are subtle and non-intrusive. All text elements are fully readable against the light background. Legibility verdict: PASS

Dark render (plot-dark.png): The same plot rendered on a warm near-black background (#1A1A17) with perfect theme adaptation. The title and axis labels now appear in light text (#F0EFE8 and #B8B7B0 respectively) and remain clearly readable against the dark background. The tick labels are equally visible in the lighter gray tone. Crucially, the scatter point color (#009E73 brand green) and LOWESS curve color (#D55E00 orange) are identical to the light render—only the chrome elements (background, text, grid) have adapted to the dark theme. Grid lines remain subtle and visible. No dark-on-dark contrast issues are present. Legibility verdict: PASS

Score: 89/100

Category	Score	Max
Visual Quality	30	30
Design Excellence	11	20
Spec Compliance	15	15
Data Quality	15	15
Code Quality	10	10
Library Mastery	8	10
Total	89	100

Visual Quality (30/30)

• VQ-01: Text Legibility (8/8) - Font sizes explicitly set (title=24pt, axis_title=20pt, axis_text=16pt). All readable in both light and dark themes.
• VQ-02: No Overlap (6/6) - Excellent spacing throughout; no overlapping text or elements.
• VQ-03: Element Visibility (6/6) - Scatter points with alpha=0.6 show density effectively; LOWESS curve is visually distinct and prominent.
• VQ-04: Color Accessibility (2/2) - Uses Okabe-Ito palette (#009E73 and #D55E00); colors are CVD-safe and provide strong contrast.
• VQ-05: Layout & Canvas (4/4) - Perfect 16:9 aspect ratio (4800×2700px); generous margins; no elements cut off.
• VQ-06: Axis Labels & Title (2/2) - Title follows format specification; axis labels descriptive with units.
• VQ-07: Palette Compliance (2/2) - First series is brand green (#009E73); accent is #D55E00 (vermillion). Plot backgrounds correct (#FAF8F1 light / #1A1A17 dark). Both renders are theme-correct.

Design Excellence (11/20)

• DE-01: Aesthetic Sophistication (4/8) - Clean, professional appearance using standard plotnine theme. No distinctive custom palette or design elements beyond the default. Competent but not sophisticated.
• DE-02: Visual Refinement (4/6) - Uses theme_minimal() appropriately; grid is subtle with proper alpha values (0.10 major, 0.05 minor); spacing is generous. Demonstrates good refinement choices.
• DE-03: Data Storytelling (3/6) - LOWESS curve creates a clear focal point against scattered data. Visual hierarchy exists but is straightforward; no exceptional narrative enhancement.

Spec Compliance (15/15)

• SC-01: Plot Type (5/5) - Correct scatter plot with LOWESS regression overlay.
• SC-02: Required Features (4/4) - Implements all requirements: scatter points with moderate transparency (alpha=0.6), smooth LOWESS curve (span=0.4 is within spec range 0.3-0.5), visually distinct (orange vs. green), smooth curve without oscillation.
• SC-03: Data Mapping (3/3) - Temperature on x-axis, crop yield on y-axis; data range covers 5-35°C and 0-80 tons/hectare; axes show all data without clipping.
• SC-04: Title & Legend (3/3) - Title: "scatter-regression-lowess · plotnine · anyplot.ai" (correct format). No legend needed for single scatter series + single regression line (appropriate omission).

Data Quality (15/15)

• DQ-01: Feature Coverage (6/6) - Shows full aspect of plot type: scatter distribution with natural variation plus fitted LOWESS trend across the temperature range.
• DQ-02: Realistic Context (5/5) - Crop yield vs. temperature is a realistic agricultural scenario; relationship is plausible (peaks around 20-25°C); no controversial or synthetic-looking data.
• DQ-03: Appropriate Scale (4/4) - Temperature range 5-35°C is realistic for growing conditions; yield values 0-80 tons/hectare are plausible for agricultural context.

Code Quality (10/10)

• CQ-01: KISS Structure (3/3) - No unnecessary functions or classes; straightforward data generation and plotting.
• CQ-02: Reproducibility (2/2) - Uses np.random.seed(42) for deterministic output.
• CQ-03: Clean Imports (2/2) - Only necessary imports; no unused modules.
• CQ-04: Code Elegance (2/2) - Proper use of geom_smooth(method='lowess', span=0.4); no simulated features or overly complex patterns.
• CQ-05: Output & API (1/1) - Saves as plot-{THEME}.png; uses current plotnine API.

Library Mastery (8/10)

• LM-01: Idiomatic Usage (5/5) - Expert ggplot2-style composition; proper use of theme() with element_rect/element_text/element_line; idiomatic geom_smooth API.
• LM-02: Distinctive Features (3/5) - Demonstrates solid mastery of plotnine's theme system with comprehensive color token customization. However, no advanced features beyond standard API usage.

Score Caps Applied

• None

Strengths

• Excellent theme adaptation: both light and dark renders are perfectly readable with theme-dependent colors correctly applied
• Proper color usage: brand green (#009E73) for scatter, complementary orange (#D55E00) for LOWESS curve
• Clean code: straightforward, reproducible, well-structured
• Specification compliance: all required features implemented correctly with appropriate parameter choices (span=0.4 within 0.3-0.5 range; alpha=0.6 within 0.5-0.7 range)
• Professional typography: explicit font sizing throughout ensures readability at high resolution

Weaknesses

• Design excellence is modest: implementation uses standard plotnine styling without distinctive creative elements
• No confidence band: specification mentions "optional" confidence band but implementation omits it (acceptable but could enhance visual richness)

Issues Found

1. DE-01 MODEST: Generic default styling without intentional design innovation
   • Fix: Consider custom color palette adjustments, intentional whitespace design, or visual emphasis techniques

AI Feedback for Next Attempt

The implementation is solid and specification-compliant. To improve design excellence, consider: adding a subtle background gradient or custom color refinements, emphasizing the LOWESS curve through visual techniques (e.g., glow effects if plotnine supports), or implementing the optional confidence band to show regression uncertainty visually.

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 of 150 points in brand green (#009E73) with alpha=0.6 transparency against a warm off-white background (#FAF8F1). A smooth orange LOWESS regression curve (#D55E00) overlays the points, clearly depicting the nonlinear relationship between temperature (x-axis, 5-35°C) and crop yield (y-axis, 0-80 tons/hectare). The title "scatter-regression-lowess · plotnine · anyplot.ai" is rendered in dark text at the top. Axis labels ("Temperature (°C)" and "Crop Yield (tons/hectare)") and tick labels are all clearly readable in dark INK tones. Subtle gridlines with low opacity provide reference without competing with data. All text elements are completely legible and properly spaced. Legibility verdict: PASS.

Dark render (plot-dark.png): The same plot rendered on a warm near-black background (#1A1A17) with all chrome (title, labels, gridlines) adapted to light colors (#F0EFE8 primary, #B8B7B0 secondary). The scatter points remain in the exact same brand green (#009E73) and the LOWESS curve in the same orange (#D55E00) as the light render—only non-data elements changed. All text is clearly visible against the dark surface with no dark-on-dark failures. Tick labels, axis labels, and title are fully readable. The greenish data colors remain prominent and well-contrasted. Legibility verdict: PASS — both themes maintain perfect readability and color consistency between renders.

Score: 97/100

Category	Score	Max
Visual Quality	30	30
Design Excellence	19	20
Spec Compliance	15	15
Data Quality	15	15
Code Quality	10	10
Library Mastery	9	10
Total	97	100

Visual Quality (30/30)

• VQ-01: Text Legibility (8/8) — All text perfectly readable in both light and dark themes
• VQ-02: No Overlap (6/6) — No text collisions or overlapping elements
• VQ-03: Element Visibility (6/6) — Scatter points and LOWESS curve clearly visible
• VQ-04: Color Accessibility (2/2) — Okabe-Ito palette is CVD-safe with adequate contrast
• VQ-05: Layout & Canvas (4/4) — Figure 16:9 (4800x2700px) with proper proportions
• VQ-06: Axis Labels & Title (2/2) — Descriptive with units; title follows spec format
• VQ-07: Palette Compliance (2/2) — Points #009E73 (BRAND), curve #D55E00, backgrounds correct for both themes

Design Excellence (19/20)

• DE-01: Aesthetic Sophistication (7/8) — Professional with intentional colors and custom theme supporting dark/light adaptation
• DE-02: Visual Refinement (6/6) — Subtle gridlines, removed top/right spines, generous whitespace
• DE-03: Data Storytelling (6/6) — LOWESS curve creates clear focal point guiding viewer to the relationship

Spec Compliance (15/15)

• SC-01: Plot Type (5/5) — Correct scatter plot with LOWESS regression
• SC-02: Required Features (4/4) — LOWESS curve visually distinct, points have moderate transparency (alpha=0.6)
• SC-03: Data Mapping (3/3) — X=temperature, Y=yield; axes show full data range
• SC-04: Title & Legend (3/3) — Title format correct; no legend needed

Data Quality (15/15)

• DQ-01: Feature Coverage (6/6) — Shows complex nonlinear agricultural relationship
• DQ-02: Realistic Context (5/5) — Crop yield vs temperature is real-world plausible with realistic peak around 22°C
• DQ-03: Appropriate Scale (4/4) — Sensible ranges for agricultural data

Code Quality (10/10)

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

Library Mastery (9/10)

• LM-01: Idiomatic Usage (5/5) — High-level API used correctly (ggplot, aes, geoms, theme)
• LM-02: Distinctive Features (4/5) — Uses geom_smooth with method='lowess' and custom theme with element_* functions

Score Caps Applied

• None

Strengths

• Perfect theme adaptation with identical data colors in both renders and proper chrome switching between light and dark
• Clean, idiomatic plotnine code using high-level API correctly throughout
• Excellent spec compliance with LOWESS regression perfectly implemented and realistic agricultural data
• Professional visual design with intentional color palette and typography hierarchy
• All text perfectly legible in both light and dark themes with zero contrast issues

Weaknesses

• Could enhance design sophistication by adding optional confidence bands around LOWESS curve (specification notes this as optional but would add distinctiveness)

Issues Found

None — implementation passes all technical requirements and visual quality standards.

AI Feedback for Next Attempt

No repair needed. This implementation is approved for production.

Verdict: APPROVED