feat(matplotlib): implement andrews-curves

[github-actions]

Implementation: andrews-curves - matplotlib

Implements the matplotlib version of andrews-curves.

File: plots/andrews-curves/implementations/matplotlib.py

Parent Issue: #2859

---

🤖 impl-generate workflow

[claude]

AI Review - Attempt 1/3

Image Description

The plot shows Andrews curves visualization using the Iris dataset with 150 observations (50 per species). Three distinct colors are used: blue (#306998) for Setosa, yellow (#FFD43B) for Versicolor, and coral/pink (#E06C75) for Virginica. The x-axis displays t values from -π to π in radians with clear labels at -π, -π/2, 0, π/2, π. The y-axis shows f(t) values ranging approximately from -4 to 5. The title "andrews-curves · matplotlib · pyplots.ai" is displayed at the top. A legend in the upper right identifies the three species. The curves clearly show Setosa (blue) clustering distinctly from the other two species, demonstrating good separation, while Versicolor (yellow) and Virginica (coral) overlap significantly but have some separation in certain regions. A subtle dashed grid (alpha=0.3) aids readability.

Quality Score: 91/100

Criteria Checklist

Visual Quality (36/40 pts)

• VQ-01: Text Legibility (10/10) - Title at 24pt, axis labels at 20pt, tick labels at 16pt - all perfectly readable
• VQ-02: No Overlap (8/8) - No overlapping text elements
• VQ-03: Element Visibility (7/8) - Lines at linewidth=1.5 with alpha=0.4 work well for 150 curves, though slightly thin per library guidelines (recommends linewidth=2-4)
• VQ-04: Color Accessibility (4/5) - Blue/yellow/coral palette is mostly colorblind-safe, though yellow can be hard to distinguish from coral for some types of colorblindness
• VQ-05: Layout Balance (5/5) - Plot fills canvas well with balanced margins
• VQ-06: Axis Labels (1/2) - "t (radians)" and "f(t)" are descriptive but f(t) could be more explanatory (e.g., "Andrews Function Value")
• VQ-07: Grid & Legend (1/2) - Grid is subtle at alpha=0.3, but legend alpha=0.8 sample lines differ from plotted alpha=0.4 curves

Spec Compliance (25/25 pts)

• SC-01: Plot Type (8/8) - Correct Andrews curves visualization
• SC-02: Data Mapping (5/5) - Variables correctly transformed using Fourier expansion
• SC-03: Required Features (5/5) - Normalized data, appropriate alpha for overlapping curves, colored by category, t from -π to π
• SC-04: Data Range (3/3) - All curves visible within axes range
• SC-05: Legend Accuracy (2/2) - Legend correctly identifies species
• SC-06: Title Format (2/2) - Uses correct format "andrews-curves · matplotlib · pyplots.ai"

Data Quality (20/20 pts)

• DQ-01: Feature Coverage (8/8) - Shows cluster separation (Setosa distinct) and overlap patterns (Versicolor/Virginica), demonstrating Andrews curves' utility for multivariate comparison
• DQ-02: Realistic Context (7/7) - Uses classic Iris dataset, a standard benchmark in data visualization
• DQ-03: Appropriate Scale (5/5) - StandardScaler normalization produces appropriate f(t) range

Code Quality (7/10 pts)

• CQ-01: KISS Structure (3/3) - Clean imports → data → plot → save structure
• CQ-02: Reproducibility (0/3) - No np.random.seed(42), though sklearn's Iris is deterministic so output is reproducible
• CQ-03: Clean Imports (2/2) - All imports used
• CQ-04: No Deprecated API (1/1) - No deprecated functions
• CQ-05: Output Correct (1/1) - Saves as plot.png

Library Features (3/5 pts)

• LF-01: Uses matplotlib's Axes methods correctly (plot, set_xlabel, etc.) but doesn't leverage advanced features like custom cyclers or matplotlib-specific color normalizations

Strengths

• Excellent implementation of Andrews curves algorithm using Fourier basis functions
• Clean visualization that clearly shows Setosa species separation from Versicolor/Virginica
• Proper normalization with StandardScaler prevents dominant variables
• Appropriate transparency (alpha=0.4) for 150 overlapping curves reveals density patterns
• X-axis uses mathematically meaningful tick labels (-π, -π/2, 0, π/2, π)

Weaknesses

• Missing np.random.seed(42) in code (though Iris dataset is deterministic, best practice is to include it)
• Linewidth=1.5 is slightly below recommended 2-4 for high-resolution plots per library guidelines
• Legend sample lines use alpha=0.8 while actual curves use alpha=0.4, creating visual inconsistency

Verdict: APPROVED