feat(bokeh): implement streamgraph-basic

[github-actions]

Implementation: streamgraph-basic - python/bokeh

Implements the python/bokeh version of streamgraph-basic.

File: plots/streamgraph-basic/implementations/python/bokeh.py

Parent Issue: #856

---

🤖 impl-generate workflow

[claude]

AI Review - Attempt 1/3

Image Description

Light render (plot-light.png): The plot displays a classic streamgraph on a warm off-white (#FAF8F1) background. Six smooth flowing bands represent music genres: Pop (brand green #009E73) at the bottom, Rock (vermillion #D55E00), Hip-Hop (blue #0072B2), Electronic (reddish purple #CC79A7), Jazz (orange #E69F00), and Classical (sky blue #56B4E9) at the top. The symmetric baseline centers the stack around y=0, with the y-axis spanning approximately -100 to +100. The title "streamgraph-basic · bokeh · anyplot.ai" is visible at top-left. Y-axis tick labels (-100, -50, 0, 50, 100) and the "Streaming Hours (relative)" axis label are clearly visible in dark ink. A legend panel sits outside the plot on the right. X-axis date labels are present at the bottom but compressed in the thumbnail. All text is readable against the light background. No light-on-light failures.

Dark render (plot-dark.png): Same streamgraph on a warm near-black (#1A1A17) background. The six data band colors are identical to the light render — Okabe-Ito positions 1–6 remain constant. All chrome (title, axis labels, tick labels, legend text) is correctly flipped to light colors (consistent with INK and INK_SOFT dark-mode tokens). The legend background renders in the dark elevated tone (#242420). Y-axis tick labels appear clearly in a warm light gray against the dark background. No dark-on-dark text failures observed. Border fill color correctly matches the plot background, creating a seamless dark canvas.

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

Score: 88/100

Category	Score	Max
Visual Quality	28	30
Design Excellence	11	20
Spec Compliance	15	15
Data Quality	15	15
Code Quality	10	10
Library Mastery	9	10
Total	88	100

Visual Quality (28/30)

• VQ-01: Text Legibility (7/8) — Font sizes explicitly set (32pt title, 24pt axis labels, 18pt ticks). All text clearly readable in both themes. Minor: x-axis date labels at the very bottom edge of the viewport are difficult to confirm in thumbnail but present in code with correct font size.
• VQ-02: No Overlap (6/6) — Streaming bands flow smoothly; no label or band collisions.
• VQ-03: Element Visibility (6/6) — All 6 bands clearly distinguishable at fill_alpha=0.85 with line borders; prominent at full resolution.
• VQ-04: Color Accessibility (2/2) — Okabe-Ito palette is CVD-safe; no red-green pairing as sole signal.
• VQ-05: Layout & Canvas (3/4) — Good 16:9 proportions; legend correctly placed outside. Minor: streams extend close to top/bottom plot edges, leaving minimal vertical breathing room.
• VQ-06: Axis Labels & Title (2/2) — Title correctly formatted as "streamgraph-basic · bokeh · anyplot.ai"; y-axis labelled "Streaming Hours (relative)"; x-axis labelled "Time".
• VQ-07: Palette Compliance (2/2) — First series (Pop) = #009E73 ✓; full Okabe-Ito order followed; light background #FAF8F1, dark background #1A1A17; both renders theme-correct.

Design Excellence (11/20)

• DE-01: Aesthetic Sophistication (5/8) — Clean professional streamgraph; Okabe-Ito colors with alpha blending, well-placed external legend, subtle grid (10% alpha), no figure outline. Raised above default for deliberate palette and layout choices.
• DE-02: Visual Refinement (3/6) — outline_line_color=None removes the figure box; grid at 10% alpha is very subtle; legend with custom glyph sizes and spacing. Raised from default.
• DE-03: Data Storytelling (3/6) — Smooth flowing curves effectively communicate genre trend dynamics; symmetric baseline creates visual balance. No annotations or callouts guiding specific insights.

Spec Compliance (15/15)

• SC-01: Plot Type (5/5) — Correct streamgraph with symmetric baseline centered around y=0; organic river-like appearance matches spec.
• SC-02: Required Features (4/4) — Smooth interpolation ✓, symmetric centered baseline ✓, distinct harmonious colors ✓, legend ✓.
• SC-03: Data Mapping (3/3) — X=datetime time axis ✓, Y=streaming hours (relative, centered) ✓, all 24 time points covered.
• SC-04: Title & Legend (3/3) — Title matches format; legend labels (Pop, Rock, Hip-Hop, Electronic, Jazz, Classical) correct.

Data Quality (15/15)

• DQ-01: Feature Coverage (6/6) — Demonstrates all key streamgraph aspects: smooth curves, symmetric baseline, 6 categories (within 3–8 spec range), 24 time points (within 10–100 range).
• DQ-02: Realistic Context (5/5) — Monthly music genre streaming data is realistic, relatable, and culturally neutral.
• DQ-03: Appropriate Scale (4/4) — Per-genre base values of 14–45 hours/month with sinusoidal variation are plausible; 2-year time range is appropriate.

Code Quality (10/10)

• CQ-01: KISS Structure (3/3) — No functions or classes; clean top-to-bottom script.
• CQ-02: Reproducibility (2/2) — np.random.seed(42) set.
• CQ-03: Clean Imports (2/2) — All imports (os, time, Path, numpy, pandas, bokeh components, selenium) are actively used.
• CQ-04: Code Elegance (2/2) — No fake UI; polynomial smoothing is functional and concise; clear section comments.
• CQ-05: Output & API (1/1) — Saves plot-{THEME}.html and plot-{THEME}.png correctly.

Library Mastery (9/10)

• LM-01: Idiomatic Usage (5/5) — ColumnDataSource per patch, figure.patch for area rendering, HoverTool via add_tools, Legend model via add_layout — all canonical bokeh patterns.
• LM-02: Distinctive Features (4/5) — Interactive HoverTool showing genre on hover, HTML artifact output, externally-placed Legend model, per-patch ColumnDataSource — distinctly Bokeh. Slightly below max as callbacks/widgets not leveraged.

Score Caps Applied

• None

Strengths

• Complete spec compliance: symmetric baseline streamgraph with smooth curves, all required features present
• Correct Okabe-Ito palette in canonical order; brand green #009E73 as first series (Pop)
• Full theme adaptation: background, borders, text, legend all correctly token-mapped for both light and dark
• Interactive HoverTool displaying genre name on hover — showcases Bokeh's distinctive interactivity
• HTML + PNG dual output following bokeh library requirements
• Proper headless Chrome screenshot via Selenium (correct approach for this environment)
• Clean, reproducible code with seed=42; no functions/classes

Weaknesses

• Polynomial fitting (degree 10 with np.polyfit/np.polyval) for smoothing rather than spline interpolation — high-degree polynomials can cause edge artifacts (Runge's phenomenon) at the leftmost and rightmost time points
• Streaming bands extend close to the top and bottom edges of the plot viewport with minimal visual margin — slight layout breathing-room issue
• No visual hierarchy or focal-point emphasis (e.g., no annotation callout for the most prominent trend, no emphasis on the largest category)

Issues Found

1. DE-03 MODERATE: No focal point or data story emphasis — Fix: Add a subtle text annotation highlighting the dominant genre or a trend crossing point to guide the viewer's eye
2. DE-01 MODERATE: Legend placed but no design hierarchy — Fix: Consider bold-facing the dominant category label or using slightly stronger alpha on the primary band to create visual emphasis

AI Feedback for Next Attempt

Implementation is strong and fully spec-compliant. For the next attempt, consider replacing degree-10 polynomial smoothing with scipy.interpolate.make_interp_spline (cubic B-spline) to avoid potential Runge edge artifacts. Add a data story element — either a subtle annotation on the most interesting trend or slight alpha variation (e.g., 0.9 for the dominant band vs 0.75 for smaller ones) to create visual hierarchy. Optionally increase bottom/top padding in the figure to give the streams more breathing room.

Verdict: APPROVED

[github-actions]

🔧 Repair Workflow Crashed (Attempt 1/4) — Auto-Retrying

The repair workflow failed (probably a transient Claude Code Action issue). Automatically re-triggering this attempt...

---

🤖 impl-repair

[github-actions]

❌ Repair Workflow Crashed (Attempt 1/4, retry exhausted)

The repair workflow itself failed twice for this attempt — likely a persistent Claude Code Action issue.

Manual restart:

gh workflow run impl-repair.yml -f pr_number=5701 -f specification_id=streamgraph-basic -f library=bokeh -f attempt=1

---

🤖 impl-repair