feat(seaborn): implement cat-strip

[github-actions]

Implementation: cat-strip - seaborn

Implements the seaborn version of cat-strip.

File: plots/cat-strip/implementations/seaborn.py

---

🤖 impl-generate workflow

[claude]

AI Review - Attempt 1/3

Image Description

The plot displays a categorical strip plot showing "Quality Score" (y-axis, range ~45-105) across five manufacturing lines (Line A through Line E on x-axis). Each category has approximately 25 data points with horizontal jitter to prevent overlap. The colors are distinct for each line: Line A uses steel blue (#306998), Line B uses golden yellow (#FFD43B), Line C uses light blue (#4B8BBE), Line D uses gray (#646464), and Line E uses bright yellow (#F0DB4F). The title "cat-strip · seaborn · pyplots.ai" appears at the top in bold. Points are well-sized with good alpha transparency (0.8). A subtle horizontal dashed grid (y-axis only) aids readability. The distributions clearly differ: Line C shows tight clustering around 90-95, Line D shows wide spread from ~55-90, Line A clusters around 80-90, Line B spreads from ~62-93, and Line E ranges from ~77-103.

Quality Score: 92/100

Criteria Checklist

Visual Quality (37/40 pts)

• VQ-01: Text Legibility (10/10) - Title 24pt bold, axis labels 20pt, tick labels 16pt - all perfectly readable
• VQ-02: No Overlap (8/8) - No text overlap, jitter prevents data point overlap
• VQ-03: Element Visibility (7/8) - Markers size=12 with alpha=0.8, good for 125 points, slightly large but visible
• VQ-04: Color Accessibility (4/5) - Colors are distinguishable, though yellow shades (Line B, E) are similar
• VQ-05: Layout Balance (5/5) - Plot fills canvas well, balanced margins with tight_layout
• VQ-06: Axis Labels (1/2) - Descriptive labels but no units ("Quality Score" could be "Quality Score (0-100)")
• VQ-07: Grid & Legend (2/2) - Subtle y-axis grid with alpha=0.3, no legend needed as x-axis labels suffice

Spec Compliance (25/25 pts)

• SC-01: Plot Type (8/8) - Correct categorical strip plot using sns.stripplot
• SC-02: Data Mapping (5/5) - Categories on x-axis, numeric values on y-axis
• SC-03: Required Features (5/5) - Jitter applied (0.25), individual points visible, categories clear
• SC-04: Data Range (3/3) - Y-axis range 45-105 shows all data with padding
• SC-05: Legend Accuracy (2/2) - No legend needed, colors match x-axis labels via hue
• SC-06: Title Format (2/2) - Correct format: "cat-strip · seaborn · pyplots.ai"

Data Quality (18/20 pts)

• DQ-01: Feature Coverage (7/8) - Shows different distributions (tight vs spread), different means, potential outliers in Line D - could show more extreme outliers
• DQ-02: Realistic Context (7/7) - Manufacturing quality scores is a plausible, neutral business scenario
• DQ-03: Appropriate Scale (4/5) - Quality scores 55-103 reasonable, though >100 is slightly odd for a "score"

Code Quality (10/10 pts)

• CQ-01: KISS Structure (3/3) - Clean imports → data → plot → save structure, no functions/classes
• CQ-02: Reproducibility (3/3) - np.random.seed(42) set
• CQ-03: Clean Imports (2/2) - All imports used (matplotlib, numpy, pandas, seaborn)
• CQ-04: No Deprecated API (1/1) - Uses modern seaborn API with hue parameter for palette
• CQ-05: Output Correct (1/1) - Saves as plot.png

Library Features (2/5 pts)

• LF-01: Uses distinctive library features (2/5) - Uses sns.stripplot correctly but doesn't leverage advanced seaborn features like combining with boxplot/violin or using built-in statistical annotations

Strengths

• Excellent use of seaborn's modern API with hue parameter to avoid deprecation warnings
• Clear visual differentiation between manufacturing lines with distinct colors
• Data demonstrates variety in distributions (tight clustering in Line C vs wide spread in Line D)
• Professional title format and appropriate font sizing throughout
• Good use of jitter (0.25) to show individual points without excessive spread

Weaknesses

• Axis labels lack units ("Quality Score" could specify scale like "(0-100)")
• Two yellow shades (Line B #FFD43B, Line E #F0DB4F) are somewhat similar - could use more distinct palette
• Does not showcase advanced seaborn features like overlaying with boxplot or using built-in palettes like 'colorblind'
• Some data points exceed 100 which is unusual for a "score" metric

Verdict: APPROVED