feat(bokeh): implement scatter-regression-lowess

[github-actions]

Implementation: scatter-regression-lowess - bokeh

Implements the bokeh version of scatter-regression-lowess.

File: plots/scatter-regression-lowess/implementations/bokeh.py

Parent Issue: #2855

---

🤖 impl-generate workflow

[claude]

AI Review - Attempt 1/3

Image Description

The plot displays a scatter plot with 200 blue data points (#306998) showing enzyme activity (%) vs temperature (°C). A smooth yellow/gold LOWESS curve (#FFD43B) is overlaid, clearly showing the characteristic bell-shaped enzyme activity pattern - activity rises from ~20% at 10°C, peaks at approximately 75% around 35°C, then decreases back to ~30% at 50°C. The title "scatter-regression-lowess · bokeh · pyplots.ai" appears in the top left. Axis labels include units (°C and %). A legend in the top right identifies "Data Points" and "LOWESS Fit". The grid uses subtle dashed lines. The overall layout is clean with good use of canvas space.

Quality Score: 92/100

Criteria Checklist

Visual Quality (36/40 pts)

• VQ-01: Text Legibility (9/10) - Title at 36pt, axis labels at 28pt, tick labels at 22pt - all clearly readable, slightly larger than minimum requirements
• VQ-02: No Overlap (8/8) - No overlapping text elements anywhere
• VQ-03: Element Visibility (7/8) - Markers (size=18) and line (width=5) are well-sized for 200 points; alpha=0.6 appropriate; minor: markers could be slightly larger
• VQ-04: Color Accessibility (5/5) - Blue (#306998) and yellow (#FFD43B) provide excellent contrast, colorblind-safe combination
• VQ-05: Layout Balance (4/5) - Good canvas utilization, plot fills majority of space; slight imbalance with legend in far corner
• VQ-06: Axis Labels (2/2) - Both axes have descriptive labels with units: "Temperature (°C)" and "Enzyme Activity (%)"
• VQ-07: Grid & Legend (1/2) - Grid subtle with alpha=0.3 and dashed style; legend positioned well but could be larger for the canvas size

Spec Compliance (25/25 pts)

• SC-01: Plot Type (8/8) - Correct: scatter plot with LOWESS regression curve
• SC-02: Data Mapping (5/5) - X=temperature (independent), Y=enzyme activity (dependent) correctly assigned
• SC-03: Required Features (5/5) - All spec features present: LOWESS curve distinct from points, moderate smoothing (frac=0.4), transparency on points, axis labels, descriptive title
• SC-04: Data Range (3/3) - Axes show all data points and full LOWESS curve
• SC-05: Legend Accuracy (2/2) - Legend correctly identifies "Data Points" and "LOWESS Fit"
• SC-06: Title Format (2/2) - Exact format: "scatter-regression-lowess · bokeh · pyplots.ai"

Data Quality (19/20 pts)

• DQ-01: Feature Coverage (8/8) - Excellent: shows complex non-linear relationship with rising, peak, and declining phases - perfect for demonstrating LOWESS capabilities
• DQ-02: Realistic Context (6/7) - Enzyme activity vs temperature is a well-known biological phenomenon; the bell curve is scientifically accurate; minor: could add more context
• DQ-03: Appropriate Scale (5/5) - Temperature range 10-50°C realistic for enzyme studies; activity 0-100% appropriate; peak at ~35°C matches typical enzyme optima

Code Quality (9/10 pts)

• CQ-01: KISS Structure (3/3) - Clean structure: imports → data → LOWESS calculation → plot → styling → save
• CQ-02: Reproducibility (3/3) - Uses np.random.seed(42)
• CQ-03: Clean Imports (2/2) - All imports are used: numpy, bokeh modules, statsmodels lowess
• CQ-04: No Deprecated API (0/1) - Using p.scatter() which is current, but CDN import for save is older pattern
• CQ-05: Output Correct (1/1) - Saves as plot.png and plot.html

Library Features (3/5 pts)

• LF-01: Uses distinctive library features (3/5) - Uses ColumnDataSource properly, figure configuration, and both PNG and HTML export; could leverage more Bokeh-specific features like HoverTool for interactivity

Strengths

• Excellent biological data example (enzyme activity curve) that perfectly demonstrates LOWESS smoothing on non-linear data
• Clean color contrast between blue points and yellow LOWESS curve makes the relationship immediately clear
• Proper use of ColumnDataSource pattern for data management
• Text sizing scaled appropriately for 4800x2700 canvas
• Correctly uses statsmodels LOWESS with appropriate smoothing fraction (0.4)
• Generates both static PNG and interactive HTML output

Weaknesses

• Legend text could be larger to match the canvas scale (currently 22pt vs 28pt axis labels)
• Could add HoverTool to show individual point values in interactive version
• Minor: grid line alpha at 0.3 is good but dashed style may be unnecessary

Verdict: APPROVED