feat(pygal): implement raincloud-basic (#1937)

## Implementation: `raincloud-basic` - pygal

Implements the **pygal** version of `raincloud-basic`.

**File:** `plots/raincloud-basic/implementations/pygal.py`

---
:robot: *[impl-generate
workflow](https://github.com/MarkusNeusinger/pyplots/actions/runs/20500987018)*

---------

Co-authored-by: github-actions[bot] <github-actions[bot]@users.noreply.github.com>
Co-authored-by: Claude Opus 4.5 <noreply@anthropic.com>
Co-authored-by: github-actions[bot] <41898282+github-actions[bot]@users.noreply.github.com>

Author: github-actions[bot]

plots/raincloud-basic/implementations/pygal.py

@@ -1,7 +1,7 @@
 """ pyplots.ai
 raincloud-basic: Basic Raincloud Plot
 Library: pygal 3.1.0 | Python 3.13.11
-Quality: 88/100 | Created: 2025-12-24
+Quality: 78/100 | Created: 2025-12-25
 """
 
 import numpy as np
@@ -22,153 +22,156 @@
 data["Treatment A"] = np.append(data["Treatment A"], [550, 200])
 data["Treatment B"] = np.append(data["Treatment B"], [480, 180])
 
-# Group colors for consistent styling
-# Pattern: cloud1, cloud2, cloud3, rain1, rain2, rain3, box elements (gray)
+# Group colors - distinct for each category (colorblind-safe)
 group_colors = ["#306998", "#FFD43B", "#4CAF50"]
-palette = tuple(group_colors) + tuple(group_colors) + ("#333333", "#333333", "#333333")
 
-# Custom style for 4800x2700 px canvas - with legend enabled
+# Custom style for 4800x2700 px canvas - scaled up for visibility
 custom_style = Style(
     background="white",
     plot_background="white",
     foreground="#333333",
     foreground_strong="#333333",
-    foreground_subtle="#666666",
-    colors=palette,  # Clouds get 0-2, rain gets 3-5 (same colors), box gets 6+
-    title_font_size=72,
-    label_font_size=48,
-    major_label_font_size=42,
-    legend_font_size=36,  # Legend for group identification
-    value_font_size=36,
-    opacity=0.6,  # Balanced opacity for visibility
+    foreground_subtle="#999999",
+    guide_stroke_color="#cccccc",
+    colors=tuple(group_colors * 3) + ("#222222",) * 30,
+    title_font_size=96,
+    label_font_size=60,
+    major_label_font_size=54,
+    legend_font_size=54,
+    value_font_size=42,
+    opacity=0.6,
     opacity_hover=0.8,
 )
 
-# Create XY chart for raincloud plot - legend enabled for group identification
+# Create VERTICAL XY chart for raincloud plot
+# X-axis = Treatment Group (category), Y-axis = Reaction Time (value)
+# For vertical orientation: cloud on RIGHT side, boxplot centered, rain on LEFT
+# This follows the spec's guidance for vertical raincloud layout
 chart = pygal.XY(
     width=4800,
     height=2700,
     style=custom_style,
     title="raincloud-basic · pygal · pyplots.ai",
     x_title="Treatment Group",
     y_title="Reaction Time (ms)",
-    show_legend=True,  # Enable legend for group identification
-    legend_at_bottom=True,  # Place legend at bottom for better layout
+    show_legend=False,
     stroke=True,
     fill=True,
     dots_size=0,
-    show_x_guides=False,
+    show_x_guides=True,
     show_y_guides=True,
     range=(100, 750),
-    xrange=(0, 5),
-    margin=50,
+    xrange=(0, 4),
+    margin=80,
+    explicit_size=True,
 )
 
-# Raincloud parameters
-violin_width = 0.35
-jitter_offset = 0.45
-box_offset = 0.05
-n_points = 80
+# Raincloud layout parameters (vertical orientation)
+# CRITICAL: Cloud on RIGHT side, boxplot centered, rain on LEFT
+# This creates the visual metaphor of rain falling from cloud
+cloud_offset = 0.28  # Cloud extends to the RIGHT (positive X offset)
+rain_offset = -0.32  # Rain falls to the LEFT (negative X offset)
+n_kde_points = 80
 
-# Pre-compute all raincloud components for each group
+# Pre-compute all raincloud components
 cloud_data = []
 rain_data = []
 box_data = []
 
 for i, (category, values) in enumerate(data.items()):
-    center_x = i + 1.5
+    center_x = i + 1  # X position for this group (1, 2, 3)
     values = np.array(values)
 
-    # --- Half-Violin (cloud) - on the left side ---
+    # --- Half-Violin (cloud) - on RIGHT side of boxplot ---
     # Compute KDE using Silverman's rule
     n = len(values)
     std = np.std(values)
-    iqr = np.percentile(values, 75) - np.percentile(values, 25)
-    bandwidth = 0.9 * min(std, iqr / 1.34) * n ** (-0.2)
+    iqr_val = np.percentile(values, 75) - np.percentile(values, 25)
+    bandwidth = 0.9 * min(std, iqr_val / 1.34) * n ** (-0.2)
 
-    # Create range of y values for density
+    # Create range of y values (reaction times) for density
     y_min, y_max = values.min(), values.max()
     padding = (y_max - y_min) * 0.1
-    y_range = np.linspace(y_min - padding, y_max + padding, n_points)
+    y_range = np.linspace(y_min - padding, y_max + padding, n_kde_points)
 
     # Gaussian kernel density estimation
     density = np.zeros_like(y_range)
     for v in values:
         density += np.exp(-0.5 * ((y_range - v) / bandwidth) ** 2)
     density /= n * bandwidth * np.sqrt(2 * np.pi)
 
-    # Normalize density to desired width
-    density = density / density.max() * violin_width
+    # Normalize density and place on RIGHT side (cloud)
+    density = density / density.max() * cloud_offset
 
-    # Create half-violin shape (only left side - the "cloud")
-    cloud_points = [(center_x - d, y) for y, d in zip(y_range, density, strict=True)]
+    # Create half-violin shape - cloud extends to right (higher X)
+    cloud_points = [(center_x + d, y) for y, d in zip(y_range, density, strict=True)]
     # Close the shape along the center line
-    cloud_points = cloud_points + [(center_x, y_range[-1]), (center_x, y_range[0]), cloud_points[0]]
-    cloud_data.append((category, cloud_points))
+    cloud_points = [(center_x, y_range[0])] + cloud_points + [(center_x, y_range[-1]), (center_x, y_range[0])]
+    cloud_data.append((category, cloud_points, group_colors[i]))
 
-    # --- Jittered Points (rain) - on the right side ---
-    np.random.seed(42 + i)  # Consistent jitter per group
+    # --- Jittered Points (rain) - on LEFT side (rain falls from cloud) ---
+    np.random.seed(42 + i)
     jitter = np.random.uniform(-0.08, 0.08, len(values))
-    rain_points = [(center_x + jitter_offset + j, float(v)) for j, v in zip(jitter, values, strict=True)]
+    rain_points = [(center_x + rain_offset + j, float(v)) for j, v in zip(jitter, values, strict=True)]
     rain_data.append((category, rain_points, group_colors[i]))
 
-    # --- Box Plot (center) ---
+    # --- Box Plot (centered at group position) ---
     median = float(np.median(values))
     q1 = float(np.percentile(values, 25))
     q3 = float(np.percentile(values, 75))
     iqr = q3 - q1
     whisker_low = float(max(values.min(), q1 - 1.5 * iqr))
     whisker_high = float(min(values.max(), q3 + 1.5 * iqr))
-    box_data.append((center_x, median, q1, q3, whisker_low, whisker_high))
+    box_data.append((center_x, median, q1, q3, whisker_low, whisker_high, group_colors[i]))
 
-# Add clouds first (series 0, 1, 2 - get colors 0, 1, 2)
-for category, cloud_points in cloud_data:
-    chart.add(category, cloud_points)
+# Add clouds (half-violins) - no labels since y-axis labels already show groups
+for _category, cloud_points, _color in cloud_data:
+    chart.add(None, cloud_points, stroke=True, fill=True)
 
-# Add rain points second (series 3, 4, 5 - get colors 3, 4, 5 = same as 0, 1, 2)
+# Add rain points - increased dots_size for visibility on 4800x2700 canvas
 for _category, rain_points, _color in rain_data:
-    chart.add(
-        None,  # No legend entry
-        rain_points,
-        stroke=False,
-        fill=False,
-        dots_size=18,  # Larger points for better visibility at full resolution
-    )
-
-# Add box plots last (no color needed, use black strokes)
-box_width = 0.15
-cap_width = 0.08
-
-for center_x, median, q1, q3, whisker_low, whisker_high in box_data:
-    # IQR box - thicker stroke for visibility
+    chart.add(None, rain_points, stroke=False, fill=False, dots_size=24)
+
+# Add box plots - vertical boxes centered at each group
+# Significantly increased line weights for 4800x2700 canvas
+box_width = 0.10
+cap_width = 0.06
+
+for center_x, median, q1, q3, whisker_low, whisker_high, _color in box_data:
+    # IQR box (vertical rectangle) - thick border for visibility
     quartile_box = [
-        (center_x + box_offset - box_width, q1),
-        (center_x + box_offset - box_width, q3),
-        (center_x + box_offset + box_width, q3),
-        (center_x + box_offset + box_width, q1),
-        (center_x + box_offset - box_width, q1),
+        (center_x - box_width, q1),
+        (center_x - box_width, q3),
+        (center_x + box_width, q3),
+        (center_x + box_width, q1),
+        (center_x - box_width, q1),
     ]
-    chart.add(None, quartile_box, stroke=True, fill=False, show_dots=False, stroke_style={"width": 8})
-
-    # Median line - significantly thicker and more prominent for distinction from quartile box
-    median_line = [(center_x + box_offset - box_width * 1.5, median), (center_x + box_offset + box_width * 1.5, median)]
-    chart.add(None, median_line, stroke=True, fill=False, show_dots=False, stroke_style={"width": 16, "dasharray": "0"})
-
-    # Whiskers - thicker lines
-    whisker_top = [(center_x + box_offset, q3), (center_x + box_offset, whisker_high)]
-    whisker_bottom = [(center_x + box_offset, q1), (center_x + box_offset, whisker_low)]
-    chart.add(None, whisker_top, stroke=True, fill=False, show_dots=False, stroke_style={"width": 8})
-    chart.add(None, whisker_bottom, stroke=True, fill=False, show_dots=False, stroke_style={"width": 8})
-
-    # Whisker caps - thicker and wider
-    cap_top = [(center_x + box_offset - cap_width, whisker_high), (center_x + box_offset + cap_width, whisker_high)]
-    cap_bottom = [(center_x + box_offset - cap_width, whisker_low), (center_x + box_offset + cap_width, whisker_low)]
-    chart.add(None, cap_top, stroke=True, fill=False, show_dots=False, stroke_style={"width": 8})
-    chart.add(None, cap_bottom, stroke=True, fill=False, show_dots=False, stroke_style={"width": 8})
-
-# X-axis labels at category positions
-chart.x_labels = ["", "Control", "Treatment A", "Treatment B", ""]
-chart.x_labels_major_count = 3
+    chart.add(None, quartile_box, stroke=True, fill=False, show_dots=False, stroke_style={"width": 20})
+
+    # Median line (horizontal line within box) - thickest for emphasis
+    median_line = [(center_x - box_width * 1.3, median), (center_x + box_width * 1.3, median)]
+    chart.add(None, median_line, stroke=True, fill=False, show_dots=False, stroke_style={"width": 28})
+
+    # Whiskers (vertical lines from box to caps)
+    whisker_bottom = [(center_x, whisker_low), (center_x, q1)]
+    whisker_top = [(center_x, q3), (center_x, whisker_high)]
+    chart.add(None, whisker_bottom, stroke=True, fill=False, show_dots=False, stroke_style={"width": 14})
+    chart.add(None, whisker_top, stroke=True, fill=False, show_dots=False, stroke_style={"width": 14})
+
+    # Whisker caps (horizontal lines at ends)
+    cap_bottom = [(center_x - cap_width, whisker_low), (center_x + cap_width, whisker_low)]
+    cap_top = [(center_x - cap_width, whisker_high), (center_x + cap_width, whisker_high)]
+    chart.add(None, cap_bottom, stroke=True, fill=False, show_dots=False, stroke_style={"width": 14})
+    chart.add(None, cap_top, stroke=True, fill=False, show_dots=False, stroke_style={"width": 14})
+
+# X-axis labels for treatment groups
+chart.x_labels = [
+    {"value": 0, "label": ""},
+    {"value": 1, "label": "Control"},
+    {"value": 2, "label": "Treatment A"},
+    {"value": 3, "label": "Treatment B"},
+    {"value": 4, "label": ""},
+]
 
 # Save outputs
 chart.render_to_file("plot.html")

plots/raincloud-basic/metadata/pygal.yaml

@@ -1,30 +1,28 @@
 library: pygal
 specification_id: raincloud-basic
-created: '2025-12-24T22:17:31Z'
-updated: '2025-12-24T22:47:58Z'
+created: '2025-12-25T07:14:50Z'
+updated: '2025-12-25T07:36:01Z'
 generated_by: claude-opus-4-5-20251101
-workflow_run: 20494684444
+workflow_run: 20500987018
 issue: 0
 python_version: 3.13.11
 library_version: 3.1.0
 preview_url: https://storage.googleapis.com/pyplots-images/plots/raincloud-basic/pygal/plot.png
 preview_thumb: https://storage.googleapis.com/pyplots-images/plots/raincloud-basic/pygal/plot_thumb.png
 preview_html: https://storage.googleapis.com/pyplots-images/plots/raincloud-basic/pygal/plot.html
-quality_score: 88
+quality_score: 78
 review:
   strengths:
-  - Excellent creative implementation of raincloud plot using pygal XY chart capabilities,
-    demonstrating how to achieve complex statistical visualizations in a library not
-    designed for them
-  - 'Properly implements all three components: half-violin density (with manual KDE),
-    jittered scatter points, and complete box plots with whiskers and caps'
-  - Realistic and meaningful data scenario (reaction times across treatment groups)
-    with appropriate outliers to showcase the visualization
-  - Good use of pygal Style system for consistent, high-resolution output with appropriate
-    font sizing
-  - Proper colorblind-friendly palette with distinct colors for each group
+  - Correct raincloud layout with cloud on right, boxplot centered, rain on left (vertical
+    orientation per spec)
+  - Excellent use of pygal XY chart to build complex composite visualization from
+    scratch
+  - Colorblind-safe palette with good contrast between groups
+  - Realistic clinical trial reaction time scenario
+  - Manual KDE implementation with proper Silverman bandwidth
+  - Proper box plot components including whisker caps
   weaknesses:
-  - Legend not rendering visibly in the output despite being configured (legend_at_bottom=True,
-    show_legend=True) - groups can only be identified via x-axis labels
-  - Box plot median lines could be more visually distinct from the quartile box edges
-    for easier statistical interpretation
+  - Legend displays despite show_legend=False setting - use chart configuration to
+    fully suppress
+  - Rain dots could be slightly larger for better visibility at 4800x2700 resolution
+  - Grid lines are present but could be more subtle (lower alpha)