feat(pygal): implement residual-plot

[github-actions]

Implementation: residual-plot - pygal

Implements the pygal version of residual-plot.

File: plots/residual-plot/implementations/pygal.py

---

🤖 impl-generate workflow

[claude]

AI Review - Attempt 1/3

Image Description

The plot displays a residual plot with fitted values on the x-axis (ranging from 10 to 100) and residuals on the y-axis (ranging from approximately -20 to 28). Blue circular markers represent normal residuals scattered around zero, while red markers highlight three outliers that exceed 2 standard deviations. A horizontal dashed gray line at y=0 serves as the zero reference. The title "residual-plot · pygal · pyplots.ai" appears at the top. A legend at the bottom shows three series: "Residuals" (blue), "Outliers (>2σ)" (red), and "Zero Reference" (gray). The plot uses a clean white background with subtle grid lines.

Quality Score: 88/100

Criteria Checklist

Visual Quality (35/40 pts)

• VQ-01: Text Legibility (9/10) - Title, axis labels, and tick marks are clearly readable at the large canvas size
• VQ-02: No Overlap (8/8) - No overlapping text elements
• VQ-03: Element Visibility (7/8) - Markers are well-sized (dots_size=18), though some points are slightly dense in the middle range
• VQ-04: Color Accessibility (5/5) - Blue and red are distinguishable; colorblind-friendly palette used
• VQ-05: Layout Balance (4/5) - Good use of canvas space, though bottom legend takes significant room
• VQ-06: Axis Labels (1/2) - Descriptive labels but missing units (e.g., could specify "Fitted Values (units)")
• VQ-07: Grid & Legend (1/2) - Grid is subtle and appropriate; legend is functional but "Zero Reference" entry is not ideal (shows a gray square instead of a line indicator)

Spec Compliance (23/25 pts)

• SC-01: Plot Type (8/8) - Correct scatter/XY chart type for residual plot
• SC-02: Data Mapping (5/5) - X = fitted values, Y = residuals - correctly assigned
• SC-03: Required Features (4/5) - Has zero reference line, outlier highlighting, appropriate data distribution; missing optional ±2σ bands mentioned in spec
• SC-04: Data Range (3/3) - All data points visible, axes show full range
• SC-05: Legend Accuracy (2/2) - Legend labels accurately describe the series
• SC-06: Title Format (2/2) - Uses correct format: "residual-plot · pygal · pyplots.ai"

Data Quality (18/20 pts)

• DQ-01: Feature Coverage (7/8) - Shows heteroscedasticity pattern (wider spread at higher fitted values), includes outliers, residuals centered around zero; demonstrates key diagnostic features well
• DQ-02: Realistic Context (6/7) - Plausible regression diagnostics scenario with realistic residual distribution
• DQ-03: Appropriate Scale (5/5) - Residual values are realistic for a regression model with fitted values in 10-100 range

Code Quality (9/10 pts)

• CQ-01: KISS Structure (3/3) - Clean structure: imports → data → plot → save, no functions/classes
• CQ-02: Reproducibility (3/3) - Uses np.random.seed(42)
• CQ-03: Clean Imports (2/2) - Only necessary imports (numpy, pygal, Style)
• CQ-04: No Deprecated API (1/1) - Uses current pygal API
• CQ-05: Output Correct (0/1) - Saves as plot.png ✓ but note: font sizes in Style are larger than library guide recommends

Library Features (3/5 pts)

• LF-01: Uses distinctive library features (3/5) - Uses XY chart, custom Style, stroke_style for dashed line, legend placement; but doesn't leverage SVG interactivity that pygal is known for

Strengths

• Excellent implementation of outlier detection and visualization with distinct coloring
• Clean separation of normal residuals vs outliers using mask
• Zero reference line implemented with dashed stroke style
• Demonstrates heteroscedasticity pattern effectively
• Font sizes appropriately scaled for 4800x2700 canvas
• Proper use of pygal's XY chart for scatter data

Weaknesses

• Missing ±2σ horizontal bands which are mentioned as optional but would enhance the diagnostic value
• Axis labels lack units (generic "Fitted Values" and "Residuals")
• Legend entry for "Zero Reference" shows a colored square instead of a line representation
• Font sizes in Style deviate significantly from library guide recommendations (title_font_size=60 vs recommended 28)

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 residual plot with blue circular markers representing normal residuals scattered across the canvas, and 3 red circular markers highlighting outliers beyond 2 standard deviations. The x-axis is labeled "Fitted Values" (showing range 10-100) and the y-axis is labeled "Residuals" (ranging from approximately -20 to +28). A subtle gray horizontal reference line at y=0 is visible. The title reads "residual-plot · pygal · pyplots.ai" in the correct format. A legend at the bottom shows three entries: "Residuals" (blue), "Outliers (>2σ)" (red), and "Zero Reference" (gray). The plot has subtle grid lines and a white background. The residuals appear randomly scattered around zero with increasing variance at higher fitted values (heteroscedasticity).

Quality Score: 78/100

Criteria Checklist

Visual Quality (32/40 pts)

• VQ-01: Text Legibility (8/10) - Title and axis labels readable but tick labels are somewhat small
• VQ-02: No Overlap (8/8) - No overlapping text elements
• VQ-03: Element Visibility (7/8) - Markers are clearly visible, good dot size, but threshold lines are not visible
• VQ-04: Color Accessibility (5/5) - Blue/red contrast is colorblind-friendly
• VQ-05: Layout Balance (3/5) - Good use of canvas but legend placement leaves white space
• VQ-06: Axis Labels (1/2) - Labels present but missing units (code specifies "$1000s" but not rendered)
• VQ-07: Grid & Legend (0/2) - Grid subtle but ±2σ threshold lines not visible in output

Spec Compliance (18/25 pts)

• SC-01: Plot Type (8/8) - Correct scatter/XY chart for residual plot
• SC-02: Data Mapping (3/5) - X-axis scale mismatch: code defines fitted values 150-500 but plot shows 10-100
• SC-03: Required Features (2/5) - Zero line present but ±2σ threshold bands not visible in rendered output
• SC-04: Data Range (3/3) - All data points visible within plot area
• SC-05: Legend Accuracy (1/2) - Legend shows "Zero Reference" but code defines "y = 0 (Perfect Fit)"
• SC-06: Title Format (2/2) - Correctly uses "residual-plot · pygal · pyplots.ai"

Data Quality (18/20 pts)

• DQ-01: Feature Coverage (8/8) - Shows heteroscedasticity, outliers, and random scatter around zero
• DQ-02: Realistic Context (5/7) - House price prediction scenario is plausible but context lost due to axis rendering
• DQ-03: Appropriate Scale (5/5) - Residual values are realistic for the context

Code Quality (10/10 pts)

• CQ-01: KISS Structure (3/3) - No functions or classes, simple linear structure
• 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) - Current pygal API
• CQ-05: Output Correct (1/1) - Saves as plot.png and plot.html

Library Features (0/5 pts)

• LF-01: Uses distinctive library features (0/5) - Basic XY chart usage, no advanced pygal features like tooltips or custom rendering

Strengths

• Proper separation of normal residuals and outliers with distinct colors
• Correct title format following pyplots.ai conventions
• Good use of custom Style for canvas sizing and fonts
• Realistic house price prediction scenario with heteroscedasticity
• Clean KISS code structure with proper reproducibility (seed=42)
• Good dot sizing for the data density (100 points)

Weaknesses

• X-axis values render incorrectly: code specifies fitted values 150-500 but plot shows 10-100 (pygal may be using index positions)
• Axis labels lose their units in rendering (should show "$1000s" suffix)
• The ±2σ threshold dashed lines are not visible in the rendered output despite being added to the chart
• Legend label mismatch: "Zero Reference" rendered vs "y = 0 (Perfect Fit)" in code

Verdict: REJECTED

[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 residual plot on a white background with blue scatter points representing normal residuals distributed around a horizontal zero reference line (dark gray/black solid line). Three red dots mark outliers exceeding ±2 standard deviations. Dashed horizontal lines at approximately ±48 indicate the ±2σ thresholds. The title "residual-plot · pygal · pyplots.ai" appears at the top center. The x-axis is labeled "Fitted Values - Predicted Price ($1000s)" ranging from 150 to 500, and the y-axis is labeled "Residuals - Actual minus Predicted ($1000s)" ranging from -100 to 100. A legend at the bottom displays five series: Residuals (blue squares), Outliers (>2σ) (red squares), Zero Reference (Perfect Fit), +2σ Threshold, and -2σ Threshold. The data shows realistic heteroscedasticity with variance increasing at higher fitted values.

Quality Score: 85/100

Criteria Checklist

Visual Quality (32/40 pts)

• VQ-01: Text Legibility (8/10) - Title and axis labels are readable but legend text at bottom is quite small
• VQ-02: No Overlap (8/8) - No overlapping text elements
• VQ-03: Element Visibility (6/8) - Markers are visible but could be slightly larger for 100 points
• VQ-04: Color Accessibility (5/5) - Blue/red color scheme is colorblind-distinguishable
• VQ-05: Layout Balance (3/5) - Good canvas utilization but legend at bottom is quite far from plot area
• VQ-06: Axis Labels (2/2) - Descriptive labels with units ($1000s)
• VQ-07: Grid & Legend (0/2) - Grid guides are good but legend items are very small and hard to read

Spec Compliance (23/25 pts)

• SC-01: Plot Type (8/8) - Correct scatter/XY plot for residuals
• SC-02: Data Mapping (5/5) - Residuals on Y-axis, fitted values on X-axis correctly
• SC-03: Required Features (5/5) - Zero line, ±2σ bands, outlier highlighting all present
• SC-04: Data Range (3/3) - All data points visible within axes range
• SC-05: Legend Accuracy (0/2) - Legend labels are accurate but text is barely legible at this scale
• SC-06: Title Format (2/2) - Correct format: residual-plot · pygal · pyplots.ai

Data Quality (18/20 pts)

• DQ-01: Feature Coverage (7/8) - Shows heteroscedasticity, outliers, and random scatter well; slight loss for limited outlier variety
• DQ-02: Realistic Context (7/7) - House price prediction scenario is plausible and relatable
• DQ-03: Appropriate Scale (4/5) - Values are realistic for house prices; residual range appropriate

Code Quality (9/10 pts)

• CQ-01: KISS Structure (3/3) - Linear structure: imports → data → plot → save
• CQ-02: Reproducibility (3/3) - np.random.seed(42) used
• CQ-03: Clean Imports (2/2) - Only necessary imports (numpy, pygal, Style)
• CQ-04: No Deprecated API (1/1) - Current pygal API used
• CQ-05: Output Correct (0/1) - Saves as plot.png but also saves plot.html (minor)

Library Features (3/5 pts)

• LF-01: Uses distinctive library features (3/5) - Good use of XY chart, custom Style, stroke_style for dashed lines, but could leverage more pygal-specific features

Strengths

• Excellent implementation of all spec requirements: zero reference line, ±2σ threshold bands, and outlier highlighting in different color
• Realistic house price prediction context with appropriate heteroscedasticity pattern
• Clean code structure following KISS principles with proper random seed
• Good use of pygal's XY chart with custom styling for high-resolution output
• Proper axis labels with units and correct title format

Weaknesses

• Legend text at bottom is too small to read clearly at the rendered scale
• Grid guides could be more subtle (lower alpha) for better visual balance

Verdict: APPROVED

This is attempt 3/3 and the score of 85 exceeds the 50-point threshold for final approval. The implementation successfully demonstrates a residual plot with all required diagnostic features.

[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 residual plot for regression diagnostics with a white background. Blue circular markers represent normal residuals (97 points) scattered around zero, while 3 red markers clearly identify outliers exceeding ±2 standard deviations. The x-axis shows "Fitted Values - Predicted Price ($1000s)" ranging from 150-500, and the y-axis displays "Residuals - Actual minus Predicted ($1000s)" ranging from -100 to 100. A solid dark horizontal line marks the zero reference (perfect fit), with dashed horizontal lines at approximately ±48 indicating the ±2σ thresholds. The title correctly follows the format "residual-plot · pygal · pyplots.ai". A legend at the bottom identifies all five series. The plot demonstrates slight heteroscedasticity with increasing variance toward higher fitted values.

Quality Score: 88/100

Criteria Checklist

Visual Quality (34/40 pts)

• VQ-01: Text Legibility (8/10) - Title and axis labels are readable but axis tick labels are somewhat small relative to the large canvas
• VQ-02: No Overlap (8/8) - No overlapping text elements
• VQ-03: Element Visibility (7/8) - Markers are visible and appropriately sized for the data density (100 points); outliers clearly distinguished in red
• VQ-04: Color Accessibility (5/5) - Blue/red color scheme is colorblind-safe with good contrast
• VQ-05: Layout Balance (4/5) - Good canvas utilization, plot area fills majority of space; slight excess whitespace at top and bottom
• VQ-06: Axis Labels (2/2) - Descriptive labels with units ($1000s)
• VQ-07: Grid & Legend (0/2) - Legend text at bottom is very small and difficult to read

Spec Compliance (23/25 pts)

• SC-01: Plot Type (8/8) - Correct scatter/XY chart for residual plot
• SC-02: Data Mapping (5/5) - Residuals on Y-axis, fitted values on X-axis (correct)
• SC-03: Required Features (5/5) - Includes zero reference line, ±2σ bands, outlier highlighting
• SC-04: Data Range (3/3) - All data points visible within axis range
• SC-05: Legend Accuracy (0/2) - Legend labels are accurate but too small to read clearly
• SC-06: Title Format (2/2) - Correct format: residual-plot · pygal · pyplots.ai

Data Quality (18/20 pts)

• DQ-01: Feature Coverage (7/8) - Shows residuals, outliers, heteroscedasticity pattern, reference lines; minor: could show more extreme outlier scenarios
• DQ-02: Realistic Context (7/7) - House price prediction is a realistic, comprehensible scenario
• DQ-03: Appropriate Scale (4/5) - Values are reasonable for house prices; residuals of ±$85K on $150-500K predictions is plausible

Code Quality (10/10 pts)

• CQ-01: KISS Structure (3/3) - Linear script with imports → data → plot → save
• CQ-02: Reproducibility (3/3) - np.random.seed(42) set
• CQ-03: Clean Imports (2/2) - Only necessary imports (numpy, pygal, Style)
• CQ-04: No Deprecated API (1/1) - Current pygal API used
• 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 pygal XY chart, custom Style, stroke_style for dashed lines, but no interactive tooltips leveraged

Strengths

• Excellent implementation of all spec requirements including zero reference line, ±2σ threshold bands, and outlier highlighting
• Clean separation of normal points (blue) and outliers (red) using color differentiation
• Realistic house price prediction scenario with appropriate heteroscedasticity pattern
• Good use of pygal's XY chart with custom styling and stroke patterns for reference lines
• Correct title format and descriptive axis labels with units

Weaknesses

• Legend text at bottom is too small to read clearly at the rendered size
• Axis tick label font size could be larger for better readability on the 4800x2700 canvas

Verdict: APPROVED

This is Attempt 3/3 and the score is 88/100 (≥50 threshold for final attempt). The implementation meets quality standards with good visual representation of a residual plot, all required features from the specification, and realistic example data. The minor issues with legend readability do not significantly impact the educational value of the plot.

[github-actions]

AI Review - Final Status

Score: 88/100 (Acceptable)

After 3 repair attempts, pygal reached 88/100.

Score ≥ 50 is acceptable for the repository. Merging.

---

🤖 impl-review