feat(bokeh): implement contour-3d

[github-actions]

Implementation: contour-3d - bokeh

Implements the bokeh version of contour-3d.

File: plots/contour-3d/implementations/bokeh.py

Parent Issue: #3230

---

🤖 impl-generate workflow

[claude]

AI Review - Attempt 1/3

Image Description

The plot displays a 3D contour surface rendered using isometric projection. The surface shows two Gaussian peaks - a primary peak near the center and a secondary smaller peak offset to the side. The surface is colored using the Viridis colormap (purple at low values through green and yellow at high values ~1.0). Dark contour lines are drawn on the surface at regular intervals, clearly delineating elevation levels. The base plane shows dashed contour projections in matching Viridis colors providing reference. Custom 3D axes are drawn with X, Y, and Z labels with arrows. A vertical colorbar on the right shows "Height (Z)" ranging from approximately 0 to 1. The title "contour-3d · bokeh · pyplots.ai" appears at top left. The background is a subtle light gray.

Quality Score: 88/100

Criteria Checklist

Visual Quality (35/40 pts)

• VQ-01: Text Legibility (9/10) - Title and axis labels are readable, colorbar text slightly small
• VQ-02: No Overlap (8/8) - No overlapping text elements
• VQ-03: Element Visibility (7/8) - Surface and contours clearly visible, quads slightly dense
• VQ-04: Color Accessibility (5/5) - Viridis is colorblind-safe
• VQ-05: Layout Balance (3/5) - Plot is well-centered but has significant empty space below
• VQ-06: Axis Labels (1/2) - Axis labels (X, Y, Z) are generic without descriptive names or units
• VQ-07: Grid & Legend (2/2) - Colorbar well-placed, subtle background grid

Spec Compliance (23/25 pts)

• SC-01: Plot Type (8/8) - Correct 3D contour plot
• SC-02: Data Mapping (5/5) - X/Y grid with Z height correctly mapped
• SC-03: Required Features (4/5) - Has surface, contour lines on surface, base plane projections, colorbar; rotation not possible in static PNG
• SC-04: Data Range (3/3) - All data visible within axes
• SC-05: Legend Accuracy (2/2) - Colorbar accurately shows height scale
• SC-06: Title Format (1/2) - Uses correct format but title font appears smaller than recommended 48pt in output

Data Quality (18/20 pts)

• DQ-01: Feature Coverage (7/8) - Shows dual Gaussian peaks with varying contour densities, but could show more complex topology
• DQ-02: Realistic Context (6/7) - Mathematical surface function, plausible but generic
• DQ-03: Appropriate Scale (5/5) - Values 0-1 are reasonable for normalized height

Code Quality (8/10 pts)

• CQ-01: KISS Structure (1/3) - Uses helper function project_3d_to_2d instead of flat script
• 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) - Modern Bokeh API used
• 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 Bokeh patches, ColorBar, Label annotations, and HTML export; however contourpy is external

Strengths

• Impressive 3D visualization using 2D Bokeh via manual isometric projection
• Proper painter's algorithm for depth sorting of surface quads
• Excellent use of contourpy for contour line generation
• Base plane contour projections as reference (matches spec suggestion)
• Well-implemented custom 3D axes with arrows
• Dual output (PNG + HTML) for flexibility

Weaknesses

• Layout has excessive empty space below the surface, reducing canvas utilization
• Helper function violates KISS principle (should be inline calculation)
• Axis labels are generic (X, Y, Z) rather than descriptive with units
• Title appears smaller in output compared to the 48pt specification

Verdict: REJECTED

[github-actions]

🔧 Repair Attempt 1/3

Applied fixes based on AI review feedback.

Status: Repair completed, re-triggering review...

---

🤖 impl-repair

[claude]

AI Review - Attempt 2/3

Image Description

The plot displays a 3D contour surface using isometric projection in Bokeh. A colored surface with two distinct Gaussian peaks is rendered—a primary peak near the center and a secondary smaller peak offset to the upper right. The surface is colored using the Viridis colormap, transitioning from deep purple (low values ~0.2) through teal/green to bright yellow (high values ~1.0). Dark contour lines are drawn directly on the surface at regular intervals, showing level curves. Three custom axis arrows (X, Y, Z) with labels are positioned at the bottom-left corner of the visualization. A vertical colorbar on the right side shows the "Height (Z)" scale. The title "contour-3d · bokeh · pyplots.ai" appears at the top-left. The base plane shows faint dashed contour projections. The background is light gray with subtle grid lines.

Quality Score: 88/100

Criteria Checklist

Visual Quality (35/40 pts)

• VQ-01: Text Legibility (8/10) - Title and colorbar text are readable; axis labels (X, Y, Z) are visible but small relative to the 4800×2700 canvas
• VQ-02: No Overlap (8/8) - No overlapping text elements
• VQ-03: Element Visibility (7/8) - Surface and contours clearly visible; base plane contours are quite faint
• VQ-04: Color Accessibility (5/5) - Viridis colormap is colorblind-safe
• VQ-05: Layout Balance (4/5) - Good use of canvas but some empty space around the plot
• VQ-06: Axis Labels (1/2) - Labels show generic "X", "Y", "Z" in the image despite more descriptive labels in code
• VQ-07: Grid & Legend (2/2) - Subtle grid, well-placed colorbar

Spec Compliance (23/25 pts)

• SC-01: Plot Type (8/8) - Correct 3D contour plot with surface and level curves
• SC-02: Data Mapping (5/5) - X, Y grid with Z height values correctly mapped
• SC-03: Required Features (4/5) - Has contour lines on surface, colorbar; base plane contours present but very faint
• SC-04: Data Range (3/3) - Full data range displayed
• SC-05: Legend Accuracy (2/2) - Colorbar accurately represents values
• SC-06: Title Format (1/2) - Uses correct format but title font could be larger

Data Quality (18/20 pts)

• DQ-01: Feature Coverage (7/8) - Shows multiple peaks with varying heights; contours demonstrate value ranges well
• DQ-02: Realistic Context (6/7) - Mathematical surface visualization is appropriate; could use more domain-specific context
• DQ-03: Appropriate Scale (5/5) - Normalized values (0-1) are appropriate for demonstration

Code Quality (9/10 pts)

• CQ-01: KISS Structure (3/3) - Linear flow: imports → data → projection → plotting → 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 (0/1) - contourpy is used which requires external dependency
• 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 Bokeh patches, lines, and ColorBar; however, does not leverage Bokeh's interactive features (callbacks, widgets) that would distinguish it from static libraries

Strengths

• Excellent implementation of 3D projection in a 2D library using painter's algorithm for depth sorting
• Clean contour line generation using contourpy with proper projection to 3D space
• Good visual appearance with Viridis colormap and well-designed custom axes with arrows
• Both PNG and HTML outputs provided for static and interactive viewing
• Reproducible with fixed random seed

Weaknesses

• Axis labels appear truncated in the rendered output—"Position X (units)" shows as just "X" in the image
• Base plane contour projections are too faint (alpha=0.4 with dashed lines makes them nearly invisible)
• Title font size could be larger for better visibility at full resolution
• Does not leverage Bokeh's interactive callback features that would make the visualization more distinctive

Verdict: APPROVED

[github-actions]

🔧 Repair Attempt 2/3

Applied fixes based on AI review feedback.

Status: Repair completed, re-triggering review...

---

🤖 impl-repair

[claude]

AI Review - Attempt 3/3

Image Description

The plot displays a 3D contour surface visualization using an isometric projection technique. The surface shows two Gaussian peaks - a primary peak near the center-back and a smaller secondary peak offset to the upper right. The surface is colored using the Viridis colormap (purple/dark blue at low values transitioning through green to yellow at peaks). Dark contour lines are drawn directly on the 3D surface at regular intervals, clearly showing the isolines. Dashed contour lines are projected onto the base plane beneath the surface for reference. Three custom axis arrows point in the X, Y, and Z directions from the back-left corner. A colorbar on the right displays "Height (Z)" ranging from approximately 0.2 to 1.0. The title "contour-3d · bokeh · pyplots.ai" appears at the top left.

Quality Score: 88/100

Criteria Checklist

Visual Quality (34/40 pts)

• VQ-01: Text Legibility (7/10) - Title and axis labels are readable but relatively small for 4800x2700 output; colorbar labels are legible
• VQ-02: No Overlap (8/8) - No overlapping text elements; all labels positioned clearly
• VQ-03: Element Visibility (8/8) - Surface quads, contour lines, and base plane contours all clearly visible with good sizing
• VQ-04: Color Accessibility (5/5) - Viridis colormap is colorblind-safe with excellent perceptual uniformity
• VQ-05: Layout Balance (3/5) - Plot is well-centered but significant white space at bottom; could use more of the canvas
• VQ-06: Axis Labels (1/2) - Labels show X, Y, Z but lack descriptive context with units
• VQ-07: Grid & Legend (2/2) - Subtle grid lines, colorbar well-placed on right

Spec Compliance (23/25 pts)

• SC-01: Plot Type (8/8) - Correct 3D contour plot with surface and isolines
• SC-02: Data Mapping (5/5) - X, Y, Z correctly assigned; proper meshgrid structure
• SC-03: Required Features (4/5) - Has surface contours, base plane projection, colorbar; rotation not applicable for static PNG
• SC-04: Data Range (3/3) - All data visible within the projected view
• SC-05: Legend Accuracy (2/2) - Colorbar correctly shows amplitude scale
• SC-06: Title Format (1/2) - Format correct but title text appears too small in final render

Data Quality (18/20 pts)

• DQ-01: Feature Coverage (7/8) - Shows multiple peaks with different heights, clear contour gradients; could show more varied terrain features
• DQ-02: Realistic Context (6/7) - Generic amplitude surface suitable for many domains; could be more specific scientific context
• DQ-03: Appropriate Scale (5/5) - Normalized 0-1 amplitude scale is appropriate; grid points (40x40) provide clear detail

Code Quality (8/10 pts)

• CQ-01: KISS Structure (2/3) - Code is linear but uses loops for projection; necessarily complex for 3D simulation in 2D library
• CQ-02: Reproducibility (3/3) - np.random.seed(42) set correctly
• CQ-03: Clean Imports (2/2) - All imports used; matplotlib only for contour extraction which is reasonable
• CQ-04: No Deprecated API (0/1) - Uses contour_set.allsegs which is deprecated in matplotlib 3.8+
• CQ-05: Output Correct (1/1) - Saves as plot.png and plot.html

Library Features (5/5 pts)

• LF-01: Uses distinctive library features - Implements HoverTool, ColorBar, LinearColorMapper, export_png, interactive HTML export with CDN resources; excellent use of Bokeh's interactive capabilities

Strengths

• Creative 3D projection implementation using painter's algorithm for depth sorting
• Excellent use of Bokeh's ColorBar, HoverTool, and HTML export features
• Clean isometric projection with custom axis arrows
• Base plane contour projection provides good reference per spec requirements
• Viridis colormap is accessible and perceptually uniform

Weaknesses

• Text elements (title, axis labels, colorbar) appear small relative to the high-resolution output
• Uses deprecated matplotlib allsegs API; should use get_paths() for future compatibility
• Axis labels lack descriptive context (just X, Y, Z instead of meaningful variable names with units)
• Layout leaves significant white space at the bottom of the canvas

Verdict: APPROVED

[github-actions]

🔧 Repair Attempt 3/3

Applied fixes based on AI review feedback.

Status: Repair completed, re-triggering review...

---

🤖 impl-repair

[claude]

AI Review - Attempt 3/3

Image Description

The plot displays a 3D contour visualization using an isometric projection technique. A surface composed of two Gaussian peaks (one primary peak near the center-left and a secondary smaller peak) is rendered using the Viridis colormap, transitioning from dark purple (low values ~0.2) to yellow (high values ~1.0). Dark contour lines are drawn on the surface showing isolines at regular intervals. The 3D coordinate system is indicated by three axis arrows (X, Y, Z) at the bottom-left corner. A vertical colorbar labeled "Height (Z)" is positioned on the right side. Dashed contour lines are projected onto the base plane for reference. The title "contour-3d · bokeh · pyplots.ai" appears in the top-left corner.

Quality Score: 88/100

Criteria Checklist

Visual Quality (35/40)

• VQ-01: Text Legibility (8/10) - Title and colorbar labels are readable, but axis labels (X, Y, Z) appear small in the rendered output compared to the code's 44pt specification
• VQ-02: No Overlap (8/8) - No overlapping elements, all text is clearly separated
• VQ-03: Element Visibility (7/8) - Surface quads and contour lines visible, but base plane contours are quite subtle
• VQ-04: Color Accessibility (5/5) - Viridis colormap is colorblind-safe, excellent choice
• VQ-05: Layout Balance (4/5) - Good 3D visualization placement, some empty space but appropriate for isometric view
• VQ-06: Axis Labels (1/2) - Axis labels show only "X", "Y", "Z" in the rendered output rather than descriptive labels with units visible in the code
• VQ-07: Grid & Legend (2/2) - Subtle grid, colorbar well-positioned

Spec Compliance (23/25)

• SC-01: Plot Type (8/8) - Correct 3D contour plot type with surface and isolines
• SC-02: Data Mapping (5/5) - X, Y grid correctly mapped to Z surface values
• SC-03: Required Features (4/5) - Has contour lines on surface, colorbar, and base plane projections; rotation not available (static Bokeh limitation)
• SC-04: Data Range (3/3) - All data visible, 40x40 grid appropriate
• SC-05: Legend Accuracy (1/2) - Colorbar shows "Height (Z)" but Z-axis label differs from what specification calls "Amplitude"
• SC-06: Title Format (2/2) - Correct format: contour-3d · bokeh · pyplots.ai

Data Quality (18/20)

• DQ-01: Feature Coverage (7/8) - Shows multiple peaks, contour levels, base projections; demonstrates 3D contour effectively
• DQ-02: Realistic Context (6/7) - Mathematical/physics-inspired surface (Gaussian peaks), generic scientific context
• DQ-03: Appropriate Scale (5/5) - Values 0-1 normalized, coordinates -3 to 3, sensible for demonstration

Code Quality (9/10)

• CQ-01: KISS Structure (2/3) - Mostly linear structure, but uses functions from matplotlib for contour extraction which adds complexity
• CQ-02: Reproducibility (3/3) - Uses np.random.seed(42) for deterministic data
• CQ-03: Clean Imports (2/2) - All imports are used
• CQ-04: No Deprecated API (1/1) - Current Bokeh API usage
• CQ-05: Output Correct (1/1) - Saves as plot.png and plot.html

Library Features (3/5)

• LF-01: Distinctive Features (3/5) - Uses Bokeh patches, ColorBar, HoverTool, and interactive toolbar; but relies on matplotlib for contour extraction rather than a pure Bokeh approach

Strengths

• Impressive manual 3D projection implementation using painter's algorithm for correct depth ordering
• Excellent use of Viridis colormap for colorblind accessibility
• Both PNG and interactive HTML output provided
• Clean contour line visualization on the surface
• Base plane contour projections add reference value
• Good use of Bokeh-specific features (ColorBar, HoverTool, interactive toolbar)

Weaknesses

• Axis labels (X, Y, Z) appear smaller than specified in code - may be a rendering/export issue
• Relies on matplotlib for contour extraction instead of a pure Bokeh solution
• Base plane dashed contours are quite subtle and hard to see
• The Z-axis label inconsistency ("Amplitude" vs "Height (Z)")

Verdict: APPROVED

[github-actions]

AI Review - Final Status

Score: 88/100 (Acceptable)

After 3 repair attempts, bokeh reached 88/100.

Score ≥ 50 is acceptable for the repository. Merging.

---

🤖 impl-review