feat(seaborn): implement rose-basic

[github-actions]

Implementation: rose-basic - python/seaborn

Implements the python/seaborn version of rose-basic.

File: plots/rose-basic/implementations/python/seaborn.py

Parent Issue: #1003

---

🤖 impl-generate workflow

[claude]

AI Review - Attempt 1/3

Image Description

Light render (plot-light.png): The plot renders on a warm off-white (#FAF8F1) background. A rose/nightingale chart displays 12 monthly rainfall bars arranged in a circular format using a sequential Blues palette (lighter shades for Jan/Feb, progressively darker shades toward Nov/Dec). All bars are clearly distinguishable with white edges providing clean separation. The title "Monthly Rainfall (mm) · rose-basic · seaborn · anyplot.ai" is bold and readable at the top. Month labels (20pt bold) are evenly distributed around the perimeter. Radial gridline labels ("100 mm", "150 mm") appear in muted gray. Value annotations (89, 72, 95, etc.) are placed near each bar tip. The "Jan" tick label at 12 o'clock position sits directly below the title text, creating slight visual crowding at the top. All text is readable against the light background — PASS. Radial labels and value annotations appear at ~14pt, below the 16pt guideline.

Dark render (plot-dark.png): The same chart renders on a near-black (#1A1A17) background. Title and all labels appear in light cream/off-white text — clearly readable. Month labels adapt correctly to the dark theme. The data bar colors (Blues sequential palette) are identical to the light render, confirming correct theme separation. Radial grid lines are visible and subtle. The darker blue bars (Oct, Nov) are still distinguishable against the dark background due to the white PAGE_BG edges. No dark-on-dark text issues detected. All text readable against the dark background — PASS. Chrome (text, grid, backgrounds) adapts correctly between themes.

Both paragraphs are required. A review that only describes one render is invalid.

Score: 80/100

Category	Score	Max
Visual Quality	25	30
Design Excellence	11	20
Spec Compliance	15	15
Data Quality	15	15
Code Quality	10	10
Library Mastery	4	10
Total	80	100

Visual Quality (25/30)

• VQ-01: Text Legibility (6/8) — Title=24pt ✓, month labels=20pt ✓, but radial gridline labels and value annotations at 14pt fall below the 16pt minimum guideline
• VQ-02: No Overlap (5/6) — Month labels clear; "Jan" tick label sits immediately below title text creating minor crowding at 12 o'clock position; no severe overlaps
• VQ-03: Element Visibility (6/6) — All 12 bars clearly visible; Blues gradient provides good differentiation; value labels aid identification
• VQ-04: Color Accessibility (2/2) — Blues sequential palette is CVD-safe; no red-green dependencies
• VQ-05: Layout & Canvas (3/4) — Correct 3600×3600 square format; good circular utilization; minor crowding near title/Jan area
• VQ-06: Axis Labels & Title (2/2) — Title includes data description and units; radial labels show "mm" units
• VQ-07: Palette Compliance (1/2) — Blues sequential palette is defensible for single-polarity data (style guide allows it), but first categorical series is not #009E73; backgrounds (#FAF8F1/#1A1A17) are correct

Design Excellence (11/20)

• DE-01: Aesthetic Sophistication (4/8) — Blues palette creates a thematic water/rainfall association; white bar edges provide clean separation; solid but within well-configured library default territory
• DE-02: Visual Refinement (4/6) — Polar spine removed; dashed grid with alpha=0.18 is subtle; bar gaps (width=0.85) add breathing room; value annotations add polish; note: style guide recommends solid grid lines, not dashed
• DE-03: Data Storytelling (3/6) — Circular arrangement inherently emphasizes June peak; sequential color provides second dimension; but no deliberate callout, annotation, or visual emphasis on the insight (peak vs. off-season contrast)

Spec Compliance (15/15)

• SC-01: Plot Type (5/5) — Correct rose/nightingale chart using polar projection; bars with radius proportional to value
• SC-02: Required Features (4/4) — Circular format, radius∝value (not area), monthly categories, 12 o'clock start, radial gridlines, consistent Blues color scheme
• SC-03: Data Mapping (3/3) — Months correctly mapped to angular positions; rainfall values mapped to radius; clockwise order from 12 o'clock
• SC-04: Title & Legend (3/3) — Title "Monthly Rainfall (mm) · rose-basic · seaborn · anyplot.ai" correct; no legend needed for single series

Data Quality (15/15)

• DQ-01: Feature Coverage (6/6) — Full 12-month seasonal cycle; value range 72–168mm shows meaningful variation; seasonal pattern (summer peak, winter trough) clearly visible
• DQ-02: Realistic Context (5/5) — Monthly rainfall is a real, neutral, scientifically relevant scenario; values are climatologically plausible
• DQ-03: Appropriate Scale (4/4) — 72–168mm range is realistic for a temperate climate; seasonal pattern direction (higher May–Aug, lower Oct–Feb) is meteorologically correct

Code Quality (10/10)

• CQ-01: KISS Structure (3/3) — Clean Imports → Data → Plot → Save structure; no functions or classes
• CQ-02: Reproducibility (2/2) — Fully deterministic hardcoded data; no randomness
• CQ-03: Clean Imports (2/2) — All four imports (os, matplotlib, numpy, seaborn) are used
• CQ-04: Code Elegance (2/2) — Clean Pythonic code; zip(..., strict=True) is a nice touch; appropriate complexity
• CQ-05: Output & API (1/1) — Saves as plot-{THEME}.png with correct dpi=300, bbox_inches='tight', facecolor=PAGE_BG

Library Mastery (4/10)

• LM-01: Idiomatic Usage (3/5) — Correctly uses sns.set_theme() for comprehensive adaptive theming and sns.color_palette() for palette generation; polar chart unavoidably falls back to matplotlib's ax.bar() — this is the correct seaborn approach for chart types not natively supported
• LM-02: Distinctive Features (1/5) — Seaborn contribution is limited to theming and palette generation; the polar chart rendering is entirely matplotlib; no seaborn-distinctive plotting function is called (e.g., no sns.barplot, sns.kdeplot, etc.)

Score Caps Applied

• None — DE-01=4 > 2 and DE-02=4 > 2, so "correct but boring" cap does not apply

Strengths

• Perfect spec compliance: correct rose chart with all required features (polar projection, radius∝value, clockwise from 12 o'clock, radial gridlines)
• Perfect code quality: clean KISS structure, deterministic data, all imports used
• Excellent data quality: realistic seasonal rainfall scenario with plausible climatological values
• Theme adaptation is correct: chrome flips between light/dark while data colors stay consistent
• Square 3600×3600 format is appropriate for radially symmetric plot
• Value annotations on each bar add useful reference information

Weaknesses

• Font sizes for radial gridline labels (fontsize=14) and value annotations (fontsize=14) are below the 16pt minimum — increase both to 16pt minimum
• Blues sequential palette does not start first categorical series with #009E73 — consider using Okabe-Ito colors for categorical treatment or use a proper continuous colormap mapped to rainfall values (not positional)
• Design sophistication is at the "well-configured default" level; needs stronger visual personality or deliberate emphasis
• No data storytelling element — the peak month (June, 168mm) could be annotated or visually highlighted to create a focal point
• Seaborn distinctive plotting features are absent; the implementation is essentially matplotlib with seaborn theming; explore whether seaborn-specific techniques (e.g., palette generation tied to actual value magnitudes) can be better leveraged
• Dashed grid lines contradict the style guide (which recommends solid thin lines); for polar charts, consider solid radial gridlines

Issues Found

1. VQ-01 PARTIAL: Radial gridline labels and value annotations at fontsize=14 — below the 16pt minimum for tick/annotation text
   • Fix: Set fontsize=16 for both ax.set_yticklabels() and the ax.text() value annotation calls
2. VQ-07 PARTIAL: Blues sequential palette used without Okabe-Ito compliance
   • Fix: Either map color to actual rainfall value using a proper continuous colormap (viridis/cividis), or use Okabe-Ito with the first bar in #009E73
3. DE-01/DE-03 LOW: Design at "well-configured default" level with no deliberate storytelling
   • Fix: Annotate the peak month (Jun 168mm) with a callout; consider using a single accent color (#009E73) to highlight the peak bar against a neutral base color for all others
4. LM-02 LOW: No seaborn-distinctive plot function used
   • Fix: Use sns.color_palette() mapped to actual data values, or consider leveraging seaborn's statistical functions for preprocessing if applicable

AI Feedback for Next Attempt

Increase all font sizes to ≥16pt minimum (radial gridline labels and value annotations are at 14pt). Address the palette: either use viridis/cividis mapped to actual rainfall values (continuous treatment) or use Okabe-Ito starting with #009E73 (categorical treatment). To improve design storytelling, highlight the peak month (Jun, 168mm) using a contrasting accent color — e.g., use a neutral gray-blue for all bars except the peak which gets #009E73 or a bold brand color. This single-emphasis technique immediately guides the viewer to the most important data point and elevates DE-01 and DE-03 significantly. Replace dashed grid with solid thin gridlines (linestyle='-') per style guide.

Verdict: REJECTED

[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 sits on a warm off-white (#FAF8F1) background — clearly not pure white, matching the required surface. Twelve equal-angle wedge bars in a Blues sequential palette (lightest blue for low-rainfall months like Feb/Oct, deepest navy for the June peak at 168 mm) radiate from the center in clockwise order starting at Jan (12 o'clock). The title "Monthly Rainfall (mm) · rose-basic · seaborn · anyplot.ai" is rendered in dark ink at 24pt bold and is fully legible. Month labels (20pt bold, dark ink) ring the perimeter; radial gridlines at 100 mm and 150 mm carry small light-gray labels with units. Numeric value labels (14pt, INK_SOFT) sit just beyond each bar tip. All text is readable against the light background with no light-on-light failures. Minor crowding: the Jan tick label falls immediately below the title, which is tight but does not collide.

Dark render (plot-dark.png): The background is warm near-black (#1A1A17) — correct. Title and all text tokens (month labels, radial labels, value labels) have flipped to light cream tones and are clearly readable against the dark surface; no dark-on-dark failures detected. The Blues bar colors are identical to the light render (only chrome flipped) — the deeper navy wedges (Jun–Jul–Aug) pop effectively against the dark background, while the lighter steel-blue wedges (Feb, Oct) remain distinguishable. Bar edges (PAGE_BG = #1A1A17 in dark mode) are invisible against the background, slightly reducing inter-bar definition, but bars are still individually identifiable by angle and color. Both renders pass the legibility check.

Score: 82/100

Category	Score	Max
Visual Quality	24	30
Design Excellence	13	20
Spec Compliance	15	15
Data Quality	15	15
Code Quality	10	10
Library Mastery	5	10
Total	82	100

Visual Quality (24/30)

• VQ-01: Text Legibility (6/8) — Title 24pt, month labels 20pt, value labels 14pt all pass. Radial tick labels at 14pt are slightly undersized for a 3600×3600 canvas.
• VQ-02: No Overlap (5/6) — Jan tick label crowds the title at the top of the polar chart but does not collide. All other labels cleanly placed.
• VQ-03: Element Visibility (5/6) — All 12 bars clearly visible. The lightest Blues bars (Feb, Oct) are less prominent on the light background but still distinguishable.
• VQ-04: Color Accessibility (2/2) — Sequential Blues palette is CVD-safe; no red-green signal used.
• VQ-05: Layout & Canvas (3/4) — 12×12 in at 300 dpi = 3600×3600 px (correct square format). Slight crowding at top between title and Jan label.
• VQ-06: Axis Labels & Title (2/2) — Descriptive title with metric and units; radial labels include "mm" unit.
• VQ-07: Palette Compliance (1/2) — Backgrounds are #FAF8F1 / #1A1A17 (correct); both renders are theme-adaptive. Blues sequential palette is justified by the spec's explicit allowance of "single color with varying saturation," but the brand-green first-series rule is not satisfied.

Design Excellence (13/20)

• DE-01: Aesthetic Sophistication (5/8) — The Blues palette serves as a dual encoder (radius + color depth both map to rainfall), which is a thoughtful design decision. Value labels on bar tips improve readability. Professional, clean look overall.
• DE-02: Visual Refinement (4/6) — 0.85 width factor creates clear inter-bar gaps; polar spine removed; dashed grid at low alpha is appropriately subtle.
• DE-03: Data Storytelling (4/6) — The seasonal pattern (summer peak Jun–Jul) is immediately legible through both the radial extension and the intensifying color. The circular arrangement reinforces the cyclical nature of monthly data.

Spec Compliance (15/15)

• SC-01: Plot Type (5/5) — Correct rose chart: equal-angle wedges with radius proportional to value.
• SC-02: Required Features (4/4) — Radial gridlines, value annotations, clockwise ordering from 12 o'clock, consistent Blues color scheme.
• SC-03: Data Mapping (3/3) — 12 equal-angle wedges for 12 months; radius proportional to rainfall; ylim correctly set to max×1.15.
• SC-04: Title & Legend (3/3) — Title follows "metric · rose-basic · seaborn · anyplot.ai" format. No legend needed for single-metric chart.

Data Quality (15/15)

• DQ-01: Feature Coverage (6/6) — Demonstrates circular arrangement, proportional radii, radial gridlines, value labels, sequential color encoding — full rose chart feature set.
• DQ-02: Realistic Context (5/5) — Monthly rainfall (mm) is a canonical, neutral, real-world use case for rose charts. Values (72–168 mm/month) are plausible for a rainy temperate region.
• DQ-03: Appropriate Scale (4/4) — 12 months (spec notes 8–12 as optimal). Rainfall range covers a meaningful spread without extreme outliers.

Code Quality (10/10)

• CQ-01: KISS Structure (3/3) — No functions or classes; flat, linear script.
• CQ-02: Reproducibility (2/2) — Hardcoded data; fully deterministic.
• CQ-03: Clean Imports (2/2) — os, matplotlib.pyplot, numpy, seaborn — all actively used.
• CQ-04: Code Elegance (2/2) — Clean angle/width calculations; no fake UI.
• CQ-05: Output & API (1/1) — Saves as plot-{THEME}.png with correct facecolor argument.

Library Mastery (5/10)

• LM-01: Idiomatic Usage (3/5) — sns.set_theme() with full rc dict is the recommended seaborn pattern for theme configuration. Actual polar bar rendering is matplotlib since seaborn has no native polar API.
• LM-02: Distinctive Features (2/5) — sns.color_palette("Blues", ...) for sequential palette generation is seaborn-native. Limited seaborn-distinctive usage beyond theme setup; core chart is matplotlib.

Score Caps Applied

• None

Strengths

• Dual encoding of rainfall through both radius and Blues color depth is an intentional and effective design choice that reinforces the data story
• Complete spec compliance: all required features (radial gridlines, value labels, clockwise-from-top orientation) are present
• Theme adaptation is fully correct in both renders — all chrome tokens flip appropriately, backgrounds match required hex values, no dark-on-dark failures
• Code quality is exemplary: flat script, deterministic data, all imports used, correct output filename

Weaknesses

• Radial tick labels at 14pt are slightly undersized for the 3600×3600 canvas; bumping to 16pt would match the style guide recommendation
• Bar edge color is set to PAGE_BG, which makes edges invisible in dark mode (dark edge on dark background) — a subtle-white or INK_SOFT edge would restore inter-bar definition in both themes
• Jan tick label crowds the title because the 12 o'clock polar tick label lands in the title pad area; increasing the title pad or adjusting label placement would remove the visual tension

Issues Found

1. VQ-01 MINOR: Radial y-tick labels at 14pt below the recommended 16pt minimum for this canvas size
   • Fix: Change fontsize to 16pt in set_yticklabels()
2. VQ-03 / DE-02 MINOR: edgecolor=PAGE_BG causes invisible edges in dark mode; bars lose definition
   • Fix: Use a semi-transparent white edge or a theme-neutral contrasting edge color
3. LM-02 LOW: Seaborn's role is limited to theme setup and palette extraction; the polar chart itself is pure matplotlib
   • Fix: Structural limitation (seaborn has no polar API) — acceptable as-is

AI Feedback for Next Attempt

The implementation is solid and fully spec-compliant. If repairing: (1) increase radial tick label fontsize from 14 to 16pt; (2) replace edgecolor=PAGE_BG with a semi-transparent white edge to maintain bar definition in dark mode; (3) consider increasing title pad from 35 to 50 to give the Jan tick label more clearance. These are polish items — the core design, data, and spec compliance are strong.

Verdict: APPROVED