feat(plotnine): implement learning-curve-basic

[github-actions]

Implementation: learning-curve-basic - plotnine

Implements the plotnine version of learning-curve-basic.

File: plots/learning-curve-basic/implementations/plotnine.py

---

🤖 impl-generate workflow

[claude]

AI Review - Attempt 1/3

Image Description

The plot shows a learning curve with two lines and confidence bands. The training score (blue, #306998) starts at ~0.88 and quickly rises to ~0.99, remaining stable. The validation score (yellow/gold, #FFD43B) starts at ~0.70 and gradually improves to ~0.89. Both lines have shaded confidence bands showing variability. The x-axis shows "Training Set Size" from approximately 50 to 800, and the y-axis shows "Accuracy Score" ranging from 0.6 to 1.0. The title follows the correct format: "learning-curve-basic · plotnine · pyplots.ai". The legend is positioned in the lower right area of the plot, clearly distinguishing "Training Score" and "Validation Score". The overall layout uses a clean minimal theme with good proportions.

Quality Score: 92/100

Criteria Checklist

Visual Quality (37/40 pts)

• VQ-01: Text Legibility (10/10) - All text is clearly readable: title is prominent, axis labels are well-sized, tick labels are legible
• VQ-02: No Overlap (8/8) - No overlapping text elements anywhere in the plot
• VQ-03: Element Visibility (8/8) - Lines are thick (size=2), confidence bands have appropriate alpha (0.25)
• VQ-04: Color Accessibility (5/5) - Blue and yellow/gold are colorblind-safe, excellent contrast
• VQ-05: Layout Balance (5/5) - Plot fills canvas well, balanced margins, legend well-positioned
• VQ-06: Axis Labels (1/2) - Descriptive labels but no units (Accuracy Score could be "Accuracy Score (0-1)")
• VQ-07: Grid & Legend (0/2) - No visible grid lines (spec mentions grid could be useful for reading values)

Spec Compliance (23/25 pts)

• SC-01: Plot Type (8/8) - Correct learning curve with training and validation lines
• SC-02: Data Mapping (5/5) - X=training set size, Y=accuracy score, correctly mapped
• SC-03: Required Features (5/5) - Shaded confidence bands, two distinct curves, clear legend
• SC-04: Data Range (3/3) - All data visible, appropriate axis ranges
• SC-05: Legend Accuracy (2/2) - Legend labels "Training Score" and "Validation Score" are accurate
• SC-06: Title Format (0/2) - Title format is correct: "learning-curve-basic · plotnine · pyplots.ai" ✓ (2/2)

Correction: SC-06 should be 2/2, adjusting total to 25/25

Data Quality (18/20 pts)

• DQ-01: Feature Coverage (7/8) - Shows typical ML learning pattern: high training score, improving validation, gap between them (variance). Could show more pronounced initial gap for clearer bias-variance illustration
• DQ-02: Realistic Context (7/7) - Realistic ML classification scenario with accuracy scores
• DQ-03: Appropriate Scale (4/5) - Values are realistic (0.7-0.99 accuracy), though starting validation at 0.70 is somewhat high for small training sets

Code Quality (10/10 pts)

• CQ-01: KISS Structure (3/3) - Clean structure: imports → data → plot → save
• CQ-02: Reproducibility (3/3) - Uses np.random.seed(42)
• CQ-03: Clean Imports (2/2) - All imports are used
• CQ-04: No Deprecated API (1/1) - Uses current plotnine API
• CQ-05: Output Correct (1/1) - Saves as plot.png

Library Features (4/5 pts)

• LF-01: Uses distinctive library features (4/5) - Good use of ggplot grammar: geom_ribbon for confidence bands, geom_line, scale_color_manual, theme customization. Could use faceting or other advanced features for bonus points

Strengths

• Excellent use of plotnine's grammar of graphics with geom_ribbon for confidence bands
• Clean, well-structured code following KISS principles
• Color scheme is visually appealing and colorblind-safe (blue/gold contrast)
• Legend positioning works well within the plot area
• Realistic ML learning curve behavior accurately depicted
• Proper title format following pyplots.ai conventions

Weaknesses

• Missing grid lines which would help read exact values from the plot
• Y-axis label could include units or range indicator (e.g., "Accuracy Score (0-1)")

Verdict: APPROVED