feat(bokeh): implement bar-3d (#3149)

## Implementation: `bar-3d` - bokeh

Implements the **bokeh** version of `bar-3d`.

**File:** `plots/bar-3d/implementations/bokeh.py`

**Parent Issue:** #2857

---
:robot: *[impl-generate
workflow](https://github.com/MarkusNeusinger/pyplots/actions/runs/20627537973)*

---------

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/bar-3d/implementations/bokeh.py

@@ -0,0 +1,301 @@
+""" pyplots.ai
+bar-3d: 3D Bar Chart
+Library: bokeh 3.8.1 | Python 3.13.11
+Quality: 91/100 | Created: 2025-12-31
+"""
+
+import numpy as np
+from bokeh.io import export_png, save
+from bokeh.models import ColorBar, Label, LinearColorMapper, Range1d
+from bokeh.palettes import Viridis256
+from bokeh.plotting import figure
+from bokeh.resources import CDN
+
+
+# Data - Quarterly sales by product category (5 products x 4 quarters)
+np.random.seed(42)
+
+products = ["Product A", "Product B", "Product C", "Product D", "Product E"]
+quarters = ["Q1", "Q2", "Q3", "Q4"]
+
+n_products = len(products)
+n_quarters = len(quarters)
+
+# Sales data (thousands of units) - varied to show different bar heights
+sales = np.array(
+    [
+        [85, 92, 78, 95],  # Product A
+        [65, 70, 88, 82],  # Product B
+        [120, 115, 130, 125],  # Product C (higher performer)
+        [45, 55, 50, 60],  # Product D (lower performer)
+        [90, 85, 95, 100],  # Product E
+    ]
+)
+
+# 3D to 2D isometric projection parameters
+elev_rad = np.radians(25)
+azim_rad = np.radians(45)
+
+# Bar dimensions in 3D space
+bar_width = 0.6
+bar_depth = 0.6
+spacing_x = 1.2  # Space between products
+spacing_y = 1.2  # Space between quarters
+
+# Normalize sales for height scaling (0-5 range for good visual proportion)
+sales_normalized = sales / sales.max() * 5
+
+# Generate all bar faces using inline projection
+all_faces = []
+z_min, z_max = 0, sales.max()
+
+for i in range(n_products):
+    for j in range(n_quarters):
+        x_center = i * spacing_x
+        y_center = j * spacing_y
+        height = sales_normalized[i, j]
+        original_value = sales[i, j]
+
+        hw = bar_width / 2
+        hd = bar_depth / 2
+
+        # Top face corners (3D)
+        top_corners_3d = [
+            (x_center - hw, y_center - hd, height),
+            (x_center + hw, y_center - hd, height),
+            (x_center + hw, y_center + hd, height),
+            (x_center - hw, y_center + hd, height),
+        ]
+        # Project top face to 2D isometric
+        top_projected = []
+        for x, y, z in top_corners_3d:
+            x_rot = x * np.cos(azim_rad) - y * np.sin(azim_rad)
+            y_rot = x * np.sin(azim_rad) + y * np.cos(azim_rad)
+            x_proj = x_rot
+            z_proj = y_rot * np.sin(elev_rad) + z * np.cos(elev_rad)
+            depth = y_rot * np.cos(elev_rad) - z * np.sin(elev_rad)
+            top_projected.append((x_proj, z_proj, depth))
+        top_xs = [p[0] for p in top_projected]
+        top_ys = [p[1] for p in top_projected]
+        avg_depth_top = np.mean([p[2] for p in top_projected])
+        all_faces.append((avg_depth_top, "top", top_xs, top_ys, original_value, height))
+
+        # Front face corners (3D)
+        front_corners_3d = [
+            (x_center + hw, y_center - hd, 0),
+            (x_center + hw, y_center + hd, 0),
+            (x_center + hw, y_center + hd, height),
+            (x_center + hw, y_center - hd, height),
+        ]
+        # Project front face to 2D isometric
+        front_projected = []
+        for x, y, z in front_corners_3d:
+            x_rot = x * np.cos(azim_rad) - y * np.sin(azim_rad)
+            y_rot = x * np.sin(azim_rad) + y * np.cos(azim_rad)
+            x_proj = x_rot
+            z_proj = y_rot * np.sin(elev_rad) + z * np.cos(elev_rad)
+            depth = y_rot * np.cos(elev_rad) - z * np.sin(elev_rad)
+            front_projected.append((x_proj, z_proj, depth))
+        front_xs = [p[0] for p in front_projected]
+        front_ys = [p[1] for p in front_projected]
+        avg_depth_front = np.mean([p[2] for p in front_projected])
+        all_faces.append((avg_depth_front, "front", front_xs, front_ys, original_value, height))
+
+        # Right side face corners (3D)
+        right_corners_3d = [
+            (x_center - hw, y_center + hd, 0),
+            (x_center + hw, y_center + hd, 0),
+            (x_center + hw, y_center + hd, height),
+            (x_center - hw, y_center + hd, height),
+        ]
+        # Project right face to 2D isometric
+        right_projected = []
+        for x, y, z in right_corners_3d:
+            x_rot = x * np.cos(azim_rad) - y * np.sin(azim_rad)
+            y_rot = x * np.sin(azim_rad) + y * np.cos(azim_rad)
+            x_proj = x_rot
+            z_proj = y_rot * np.sin(elev_rad) + z * np.cos(elev_rad)
+            depth = y_rot * np.cos(elev_rad) - z * np.sin(elev_rad)
+            right_projected.append((x_proj, z_proj, depth))
+        right_xs = [p[0] for p in right_projected]
+        right_ys = [p[1] for p in right_projected]
+        avg_depth_right = np.mean([p[2] for p in right_projected])
+        all_faces.append((avg_depth_right, "right", right_xs, right_ys, original_value, height))
+
+# Sort by depth (back to front - painter's algorithm)
+all_faces.sort(key=lambda f: f[0], reverse=True)
+
+# Color mapping using Viridis colormap based on original sales values
+color_mapper = LinearColorMapper(palette=Viridis256, low=z_min, high=z_max)
+
+# Create Bokeh figure
+p = figure(
+    width=4800,
+    height=2700,
+    title="bar-3d · bokeh · pyplots.ai",
+    toolbar_location="right",
+    tools="pan,wheel_zoom,box_zoom,reset,save",
+)
+
+# Draw bar faces with shading and semi-transparency to reveal hidden bars
+for _depth, face_type, xs, ys, value, _h in all_faces:
+    # Get base color from Viridis colormap
+    idx = int((value - z_min) / (z_max - z_min) * 255)
+    idx = max(0, min(255, idx))
+    base_color = Viridis256[idx]
+
+    # Convert hex to RGB for shading
+    r = int(base_color[1:3], 16)
+    g = int(base_color[3:5], 16)
+    b = int(base_color[5:7], 16)
+
+    # Apply shading: top is brightest, front slightly darker, right side darkest
+    if face_type == "top":
+        factor = 1.0
+    elif face_type == "front":
+        factor = 0.85
+    else:  # right
+        factor = 0.7
+
+    r = int(min(255, r * factor))
+    g = int(min(255, g * factor))
+    b = int(min(255, b * factor))
+
+    color = f"#{r:02x}{g:02x}{b:02x}"
+
+    # Semi-transparent bars (alpha=0.75) to reveal hidden bars behind taller ones
+    p.patch(x=xs, y=ys, fill_color=color, line_color="#306998", line_width=1.5, line_alpha=0.6, alpha=0.75)
+
+# Calculate plot range from all face coordinates
+all_xs = [x for f in all_faces for x in f[2]]
+all_ys = [y for f in all_faces for y in f[3]]
+
+x_min, x_max = min(all_xs), max(all_xs)
+y_min, y_max = min(all_ys), max(all_ys)
+
+x_pad = (x_max - x_min) * 0.18
+y_pad = (y_max - y_min) * 0.15
+
+p.x_range = Range1d(x_min - x_pad * 1.5, x_max + x_pad * 1.5)
+p.y_range = Range1d(y_min - y_pad * 1.5, y_max + y_pad)
+
+# Hide default axes for cleaner 3D projection look
+p.xaxis.visible = False
+p.yaxis.visible = False
+
+# Add 3D axis lines from origin - inline projection
+origin_3d = (-0.5, -0.5, 0)
+origin_x_rot = origin_3d[0] * np.cos(azim_rad) - origin_3d[1] * np.sin(azim_rad)
+origin_y_rot = origin_3d[0] * np.sin(azim_rad) + origin_3d[1] * np.cos(azim_rad)
+origin_x = origin_x_rot
+origin_y = origin_y_rot * np.sin(elev_rad) + origin_3d[2] * np.cos(elev_rad)
+
+axis_color = "#444444"
+axis_width = 4
+
+# X-axis (products direction) - inline projection
+x_axis_end_3d = ((n_products - 1) * spacing_x + 0.8, -0.5, 0)
+x_axis_x_rot = x_axis_end_3d[0] * np.cos(azim_rad) - x_axis_end_3d[1] * np.sin(azim_rad)
+x_axis_y_rot = x_axis_end_3d[0] * np.sin(azim_rad) + x_axis_end_3d[1] * np.cos(azim_rad)
+x_axis_end_x = x_axis_x_rot
+x_axis_end_y = x_axis_y_rot * np.sin(elev_rad) + x_axis_end_3d[2] * np.cos(elev_rad)
+p.line(x=[origin_x, x_axis_end_x], y=[origin_y, x_axis_end_y], line_color=axis_color, line_width=axis_width)
+
+# Y-axis (quarters direction) - inline projection
+y_axis_end_3d = (-0.5, (n_quarters - 1) * spacing_y + 0.8, 0)
+y_axis_x_rot = y_axis_end_3d[0] * np.cos(azim_rad) - y_axis_end_3d[1] * np.sin(azim_rad)
+y_axis_y_rot = y_axis_end_3d[0] * np.sin(azim_rad) + y_axis_end_3d[1] * np.cos(azim_rad)
+y_axis_end_x = y_axis_x_rot
+y_axis_end_y = y_axis_y_rot * np.sin(elev_rad) + y_axis_end_3d[2] * np.cos(elev_rad)
+p.line(x=[origin_x, y_axis_end_x], y=[origin_y, y_axis_end_y], line_color=axis_color, line_width=axis_width)
+
+# Z-axis (height direction) - inline projection
+z_axis_end_3d = (-0.5, -0.5, 5.8)
+z_axis_x_rot = z_axis_end_3d[0] * np.cos(azim_rad) - z_axis_end_3d[1] * np.sin(azim_rad)
+z_axis_y_rot = z_axis_end_3d[0] * np.sin(azim_rad) + z_axis_end_3d[1] * np.cos(azim_rad)
+z_axis_end_x = z_axis_x_rot
+z_axis_end_y = z_axis_y_rot * np.sin(elev_rad) + z_axis_end_3d[2] * np.cos(elev_rad)
+p.line(x=[origin_x, z_axis_end_x], y=[origin_y, z_axis_end_y], line_color=axis_color, line_width=axis_width)
+
+# Add product labels along X-axis
+for i, product in enumerate(products):
+    label_pos_3d = (i * spacing_x, -1.0, 0)
+    lx_rot = label_pos_3d[0] * np.cos(azim_rad) - label_pos_3d[1] * np.sin(azim_rad)
+    ly_rot = label_pos_3d[0] * np.sin(azim_rad) + label_pos_3d[1] * np.cos(azim_rad)
+    label_x = lx_rot
+    label_y = ly_rot * np.sin(elev_rad) + label_pos_3d[2] * np.cos(elev_rad)
+    label = Label(
+        x=label_x, y=label_y - 0.3, text=product, text_font_size="26pt", text_color="#333333", text_align="center"
+    )
+    p.add_layout(label)
+
+# Add quarter labels along Y-axis
+for j, quarter in enumerate(quarters):
+    label_pos_3d = (-1.0, j * spacing_y, 0)
+    lx_rot = label_pos_3d[0] * np.cos(azim_rad) - label_pos_3d[1] * np.sin(azim_rad)
+    ly_rot = label_pos_3d[0] * np.sin(azim_rad) + label_pos_3d[1] * np.cos(azim_rad)
+    label_x = lx_rot
+    label_y = ly_rot * np.sin(elev_rad) + label_pos_3d[2] * np.cos(elev_rad)
+    label = Label(
+        x=label_x - 0.4, y=label_y, text=quarter, text_font_size="26pt", text_color="#333333", text_align="right"
+    )
+    p.add_layout(label)
+
+# Add axis titles
+products_label = Label(
+    x=x_axis_end_x + 0.3,
+    y=x_axis_end_y - 0.4,
+    text="Products",
+    text_font_size="32pt",
+    text_color="#333333",
+    text_font_style="bold",
+)
+p.add_layout(products_label)
+
+quarters_label = Label(
+    x=y_axis_end_x - 1.0,
+    y=y_axis_end_y - 0.3,
+    text="Quarters",
+    text_font_size="32pt",
+    text_color="#333333",
+    text_font_style="bold",
+)
+p.add_layout(quarters_label)
+
+# Add color bar for sales value scale
+color_bar = ColorBar(
+    color_mapper=color_mapper,
+    width=60,
+    location=(0, 0),
+    title="Sales (thousands)",
+    title_text_font_size="28pt",
+    major_label_text_font_size="22pt",
+    title_standoff=20,
+    margin=40,
+    padding=20,
+)
+p.add_layout(color_bar, "right")
+
+# Title styling for large canvas
+p.title.text_font_size = "44pt"
+p.title.text_font_style = "bold"
+
+# Grid styling - subtle
+p.xgrid.grid_line_color = "#dddddd"
+p.ygrid.grid_line_color = "#dddddd"
+p.xgrid.grid_line_alpha = 0.2
+p.ygrid.grid_line_alpha = 0.2
+p.xgrid.grid_line_dash = [6, 4]
+p.ygrid.grid_line_dash = [6, 4]
+
+# Background styling
+p.background_fill_color = "#f9f9f9"
+p.border_fill_color = "white"
+p.outline_line_color = None
+p.min_border_right = 220
+
+# Save PNG
+export_png(p, filename="plot.png")
+
+# Save HTML for interactive version
+save(p, filename="plot.html", resources=CDN, title="bar-3d · bokeh · pyplots.ai")

plots/bar-3d/metadata/bokeh.yaml

@@ -0,0 +1,25 @@
+library: bokeh
+specification_id: bar-3d
+created: '2025-12-31T21:36:45Z'
+updated: '2025-12-31T21:45:07Z'
+generated_by: claude-opus-4-5-20251101
+workflow_run: 20627537973
+issue: 2857
+python_version: 3.13.11
+library_version: 3.8.1
+preview_url: https://storage.googleapis.com/pyplots-images/plots/bar-3d/bokeh/plot.png
+preview_thumb: https://storage.googleapis.com/pyplots-images/plots/bar-3d/bokeh/plot_thumb.png
+preview_html: https://storage.googleapis.com/pyplots-images/plots/bar-3d/bokeh/plot.html
+quality_score: 91
+review:
+  strengths:
+  - Excellent 3D isometric projection with proper painter's algorithm for correct
+    face ordering
+  - Semi-transparency (alpha=0.75) effectively reveals occluded bars as spec recommends
+  - Viridis color gradient reinforces z-values for better depth perception per spec
+  - Face shading (top=100%, front=85%, right=70%) provides convincing 3D appearance
+  - Clean axis labels positioned correctly in 3D space
+  - Both PNG and interactive HTML outputs provided
+  - Realistic business scenario with meaningful variation in data
+  weaknesses:
+  - Title format uses hyphens/dashes instead of middle dots as required by SC-06