feat(altair): implement bar-3d

[github-actions]

Implementation: bar-3d - altair

Implements the altair version of bar-3d.

File: plots/bar-3d/implementations/altair.py

Parent Issue: #2857

---

🤖 impl-generate workflow

[claude]

AI Review - Attempt 1/3

Image Description

The plot displays a simulated 3D bar chart using isometric projection in Altair. Four products (A, B, C, D) are shown on the x-axis with bars stacked in rows representing quarters (Q1-Q4). Each bar consists of three faces: front, top, and right side, with varying brightness to create depth. Colors use the viridis colormap ranging from dark purple/blue (low sales ~56K) to bright yellow (high sales ~200K). Product B has the tallest bar (bright yellow, Q4). The title "bar-3d · altair · pyplots.ai" appears at the top with a subtitle explaining the isometric projection. A legend on the right shows the Sales ($K) color scale. Quarter labels (Q1-Q4) are positioned on the right side of the chart.

Quality Score: 86/100

Criteria Checklist

Visual Quality (34/40 pts)

• VQ-01: Text Legibility (9/10) - Title, axis labels, and product labels are clearly readable. Quarter labels slightly small but acceptable.
• VQ-02: No Overlap (7/8) - Minor overlap between some bar faces due to the 3D simulation, but overall readable.
• VQ-03: Element Visibility (7/8) - Bars are clearly visible with good sizing. The 3D effect using front/top/side faces works well.
• VQ-04: Color Accessibility (5/5) - Uses viridis colormap which is colorblind-safe.
• VQ-05: Layout Balance (4/5) - Good use of canvas space, though some empty area on the left side.
• VQ-06: Axis Labels (1/2) - Y-axis has "Sales Revenue (Relative Height)" which is descriptive but lacks actual units.
• VQ-07: Grid & Legend (1/2) - Grid is subtle (alpha 0.2), legend well placed. However, grid lines don't align well with the 3D perspective.

Spec Compliance (22/25 pts)

• SC-01: Plot Type (8/8) - Correctly implements a 3D bar chart using isometric projection.
• SC-02: Data Mapping (5/5) - Products on x-axis, quarters as depth dimension, sales as bar height.
• SC-03: Required Features (4/5) - Has 3D effect with color gradients, but lacks true interactive rotation (spec notes it as beneficial).
• SC-04: Data Range (3/3) - All data visible, axes show complete range.
• SC-05: Legend Accuracy (2/2) - Legend correctly shows Sales ($K) with viridis scale.
• SC-06: Title Format (2/2) - Uses correct format: "bar-3d · altair · pyplots.ai"

Data Quality (18/20 pts)

• DQ-01: Feature Coverage (7/8) - Shows variation across products and quarters with seasonal trends. Good variety in bar heights.
• DQ-02: Realistic Context (7/7) - Quarterly sales by product is a realistic business scenario.
• DQ-03: Appropriate Scale (4/5) - Sales values (56K-200K) are reasonable, though the relative height scale on y-axis is abstract.

Code Quality (9/10 pts)

• CQ-01: KISS Structure (3/3) - Linear flow: imports → data → plot → save. No functions/classes.
• CQ-02: Reproducibility (3/3) - Uses np.random.seed(42).
• CQ-03: Clean Imports (2/2) - Only uses altair, numpy, pandas - all necessary.
• CQ-04: No Deprecated API (1/1) - Uses current Altair API.
• CQ-05: Output Correct (0/1) - Saves both plot.png and plot.html, but minor: saves HTML which is good for Altair.

Library Features (3/5 pts)

• LF-01: Uses distinctive library features (3/5) - Uses mark_rect with layering, tooltips, and proper encoding. However, doesn't use Altair's interactive features (zoom/pan) which would enhance the visualization.

Strengths

• Creative solution to simulate 3D in a 2D library using isometric projection with multiple rect faces per bar
• Excellent use of viridis colormap with opacity variation for shading effect
• Well-structured painter's algorithm approach (sorting by depth) ensures proper occlusion
• Good tooltip implementation for interactivity in the HTML output
• Clear labeling with product names and quarter indicators

Weaknesses

• The 3D simulation, while creative, results in some visual artifacts where bar faces don't perfectly align
• Y-axis shows "Relative Height" which is abstract - would benefit from showing actual sales values
• Quarter labels positioned on the right could be more integrated into the 3D perspective
• Missing Altair's .interactive() feature for zoom/pan which would allow better exploration of the dense visualization

Verdict: REJECTED

The implementation is creative and demonstrates good understanding of both Altair and 3D visualization concepts. However, the score of 86 falls below the 90-point threshold. The main issues are the abstract y-axis labeling and some visual alignment artifacts in the 3D simulation. The implementation would benefit from better axis labeling and potentially using Altair's interactive features.

[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 isometric bar chart showing quarterly sales data across four products (Product A, B, C, D). Each bar is rendered with three faces (front, top, side) to create a 3D effect, using the viridis color scale to encode sales values from ~56K to 200K. The bars are arranged in a grid with products on the x-axis and quarters (Q1-Q4) progressing into the depth dimension. The title reads "bar-3d · altair · pyplots.ai" with a subtitle "Quarterly Sales by Product (Isometric 3D Projection)". A color legend for "Sales ($K)" is positioned on the right. Quarter labels (Q1-Q4) are visible on the right side of the chart, and an angled "Quarters →" indicator is shown. Product B has the highest bar (yellow, ~200K), while Product C shows the lowest values (purple/blue tones).

Quality Score: 91/100

Criteria Checklist

Visual Quality (35/40 pts)

• VQ-01: Text Legibility (9/10) - Title and labels are clearly readable; tick labels slightly dense
• VQ-02: No Overlap (8/8) - No overlapping text elements
• VQ-03: Element Visibility (7/8) - 3D bar faces are visible; some occlusion but depth ordering is correct
• VQ-04: Color Accessibility (5/5) - Viridis colormap is colorblind-safe
• VQ-05: Layout Balance (4/5) - Good use of canvas; slight imbalance with more space on left
• VQ-06: Axis Labels (1/2) - Y-axis shows "Sales Revenue (Relative Height)" but the actual scale values don't map directly to sales amounts
• VQ-07: Grid & Legend (1/2) - Grid is subtle; legend well placed

Spec Compliance (23/25 pts)

• SC-01: Plot Type (8/8) - Correctly implements 3D bar chart using isometric projection
• SC-02: Data Mapping (5/5) - X (products), Y (quarters as depth), Z (sales as height) correctly mapped
• SC-03: Required Features (4/5) - Has color gradient for z-values; interactivity enabled; minor: no transparency for occluded bars
• SC-04: Data Range (3/3) - All data visible
• SC-05: Legend Accuracy (2/2) - Legend correctly shows sales scale
• SC-06: Title Format (1/2) - Title format correct; subtitle adds context but y-axis label is misleading

Data Quality (20/20 pts)

• DQ-01: Feature Coverage (8/8) - Shows variation across products and quarters; seasonal patterns and growth trends visible
• DQ-02: Realistic Context (7/7) - Quarterly sales by product is a realistic business scenario
• DQ-03: Appropriate Scale (5/5) - Sales values ($56K-$200K) are realistic

Code Quality (10/10 pts)

• CQ-01: KISS Structure (3/3) - Linear script structure with imports → data → plot → save
• CQ-02: Reproducibility (3/3) - Uses np.random.seed(42)
• CQ-03: Clean Imports (2/2) - Only necessary imports
• CQ-04: No Deprecated API (1/1) - Uses current Altair API
• 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) - Good use of mark_rect, tooltips, and interactive(); could better leverage Altair's declarative grammar for the layering

Strengths

• Excellent creative solution using isometric projection to simulate 3D in a 2D library
• Good painter's algorithm implementation for proper depth ordering of bar faces
• Viridis colormap provides clear value encoding and is colorblind-accessible
• Interactive tooltips show exact product, quarter, and sales values
• Realistic business data with seasonal variation and growth trends
• Clean code structure following KISS principles

Weaknesses

• Y-axis label says "Sales Revenue (Relative Height)" but the axis tick values don't correspond to actual sales values, causing confusion
• Quarter labels positioning could be clearer - they overlap with the depth axis indicator
• The spec suggests semi-transparent bars for revealing occluded bars, which is not implemented

Verdict: APPROVED

The implementation successfully creates a 3D bar chart effect using Altair's 2D primitives with an isometric projection technique. The visualization is creative, the data is realistic, and the code quality is high. The minor issues with axis labeling and lack of transparency for occlusion don't significantly detract from the overall quality of this publication-ready visualization.