Merge branch 'main' into implementation/violin-basic/bokeh

Author: github-actions[bot]

plots/violin-basic/implementations/altair.py

@@ -1,7 +1,7 @@
 """ pyplots.ai
 violin-basic: Basic Violin Plot
-Library: altair 6.0.0 | Python 3.13.11
-Quality: 92/100 | Created: 2025-12-23
+Library: altair 6.0.0 | Python 3.14.3
+Quality: 94/100 | Updated: 2026-02-21
 """
 
 import altair as alt
@@ -16,39 +16,29 @@
 
 for cat in categories:
     if cat == "Engineering":
-        # Higher salaries with moderate spread
         values = np.random.normal(92000, 16000, 150)
     elif cat == "Marketing":
-        # Mid-range salaries
         values = np.random.normal(70000, 13000, 150)
     elif cat == "Sales":
         # Bimodal: base salary + high performers with commissions
         values = np.concatenate([np.random.normal(50000, 8000, 75), np.random.normal(92000, 11000, 75)])
     else:  # Support
-        # Lower salary, tighter distribution
         values = np.random.normal(55000, 10000, 150)
 
     for v in values:
         data.append({"Department": cat, "Salary": v})
 
 df = pd.DataFrame(data)
 
-# Calculate statistics for quartile markers
-stats = (
-    df.groupby("Department")["Salary"]
-    .agg(q1=lambda x: x.quantile(0.25), median=lambda x: x.quantile(0.5), q3=lambda x: x.quantile(0.75))
-    .reset_index()
-)
-
-# Merge stats for layering
-df_with_stats = df.merge(stats, on="Department")
+# Department order: unimodal distributions first, bimodal Sales last as focal point
+dept_order = ["Support", "Marketing", "Engineering", "Sales"]
 
-# Colors - Python palette
-colors = ["#306998", "#FFD43B", "#4B8BBE", "#FFE873"]
-color_scale = alt.Scale(domain=categories, range=colors)
+# Colors - four fully distinct colorblind-safe hues with Python Blue
+# brown, purple, Python Blue, orange — each maximally distinct
+palette = ["#8B6C42", "#9467BD", "#306998", "#E5832D"]
+color_scale = alt.Scale(domain=dept_order, range=palette)
 
-# Base chart
-base = alt.Chart(df_with_stats)
+base = alt.Chart(df)
 
 # Violin shape using kernel density transform
 violin = (
@@ -69,14 +59,25 @@
             axis=alt.Axis(labels=False, values=[0], grid=False, ticks=False),
         ),
         color=alt.Color("Department:N", scale=color_scale, legend=None),
+        tooltip=[alt.Tooltip("Department:N"), alt.Tooltip("Salary:Q", format="$,.0f")],
     )
 )
 
-# IQR rule (black vertical line)
-quartile_rule = base.mark_rule(color="black", strokeWidth=5).encode(y="q1:Q", y2="q3:Q")
+# IQR rule via declarative aggregate (one rule per department)
+quartile_rule = (
+    base.transform_aggregate(q1="q1(Salary)", q3="q3(Salary)", groupby=["Department"])
+    .mark_rule(color="#1a1a1a", strokeWidth=5)
+    .encode(y="q1:Q", y2="q3:Q")
+)
 
-# Median point (white dot with black border)
-median_point = base.mark_point(color="white", size=250, filled=True, strokeWidth=3, stroke="black").encode(y="median:Q")
+# Median point via declarative aggregate (one dot per department)
+median_point = (
+    base.transform_aggregate(med="median(Salary)", groupby=["Department"])
+    .mark_point(color="white", size=250, filled=True, strokeWidth=3, stroke="#1a1a1a")
+    .encode(
+        y="med:Q", tooltip=[alt.Tooltip("Department:N"), alt.Tooltip("med:Q", title="Median Salary", format="$,.0f")]
+    )
+)
 
 # Combine layers and facet by department
 chart = (
@@ -85,14 +86,14 @@
         column=alt.Column(
             "Department:N",
             header=alt.Header(labelFontSize=20, labelOrient="bottom", title=None, labelPadding=15),
-            sort=categories,
+            sort=dept_order,
         )
     )
     .resolve_scale(x="independent")
     .properties(title=alt.Title("violin-basic · altair · pyplots.ai", fontSize=28, anchor="middle"))
     .configure_facet(spacing=20)
     .configure_view(stroke=None, continuousWidth=350, continuousHeight=750)
-    .configure_axis(labelFontSize=18, titleFontSize=22, gridOpacity=0.3, gridDash=[3, 3])
+    .configure_axis(labelFontSize=18, titleFontSize=22, gridOpacity=0.2, gridDash=[3, 3])
 )
 
 # Save outputs

plots/violin-basic/implementations/letsplot.py

@@ -1,7 +1,7 @@
 """ pyplots.ai
 violin-basic: Basic Violin Plot
-Library: letsplot 4.8.1 | Python 3.13.11
-Quality: 91/100 | Created: 2025-12-23
+Library: letsplot 4.8.2 | Python 3.14.3
+Quality: 90/100 | Updated: 2026-02-21
 """
 
 import numpy as np
@@ -15,53 +15,66 @@
 # Data
 np.random.seed(42)
 
-categories = ["Engineering", "Marketing", "Sales", "Design"]
-colors = ["#306998", "#FFD43B", "#4B8BBE", "#FFE873"]
+# Ordered by median salary (high → low) for visual storytelling
+dept_order = ["Engineering", "Design", "Marketing", "Sales"]
+palette = ["#306998", "#2E8B57", "#E8A317", "#E07A5F"]
 
-# Generate realistic salary distributions per department
 data = []
-distributions = {
-    "Engineering": {"mean": 95000, "std": 20000, "n": 200},
-    "Marketing": {"mean": 75000, "std": 15000, "n": 150},
-    "Sales": {"mean": 70000, "std": 25000, "n": 180},
-    "Design": {"mean": 80000, "std": 18000, "n": 120},
-}
 
-for cat in categories:
-    dist = distributions[cat]
-    values = np.random.normal(dist["mean"], dist["std"], dist["n"])
-    values = np.clip(values, 30000, 200000)  # Realistic salary bounds
-    for v in values:
-        data.append({"Department": cat, "Salary": v})
+# Engineering: bimodal (junior ~$70k + senior ~$115k) — showcases violin strength
+eng_junior = np.random.normal(70000, 8000, 80)
+eng_senior = np.random.normal(115000, 12000, 120)
+eng_values = np.clip(np.concatenate([eng_junior, eng_senior]), 30000, 200000)
+for v in eng_values:
+    data.append({"Department": "Engineering", "Salary": v})
+
+# Design: moderate spread, roughly normal
+design_values = np.random.normal(80000, 18000, 120)
+design_values = np.clip(design_values, 30000, 200000)
+for v in design_values:
+    data.append({"Department": "Design", "Salary": v})
+
+# Marketing: narrower with a small cluster of high earners
+mkt_base = np.random.normal(72000, 12000, 130)
+mkt_high = np.random.normal(105000, 8000, 20)
+mkt_values = np.clip(np.concatenate([mkt_base, mkt_high]), 30000, 200000)
+for v in mkt_values:
+    data.append({"Department": "Marketing", "Salary": v})
+
+# Sales: right-skewed (many moderate earners, few top performers)
+sales_values = np.random.exponential(20000, 180) + 45000
+sales_values = np.clip(sales_values, 30000, 200000)
+for v in sales_values:
+    data.append({"Department": "Sales", "Salary": v})
 
 df = pd.DataFrame(data)
 
 # Plot
 plot = (
     ggplot(df, aes(x="Department", y="Salary", fill="Department"))  # noqa: F405
     + geom_violin(  # noqa: F405
-        quantiles=[0.25, 0.5, 0.75],  # Show quartiles including median
-        quantile_lines=True,  # Draw lines at quantiles
-        size=1.5,  # Border thickness
-        alpha=0.8,
-        trim=False,  # Show full tails
+        quantiles=[0.25, 0.5, 0.75], quantile_lines=True, size=1.2, alpha=0.85, trim=False, color="#2C3E50"
+    )
+    + scale_x_discrete(limits=dept_order)  # noqa: F405
+    + scale_fill_manual(values=dict(zip(dept_order, palette, strict=True)))  # noqa: F405
+    + scale_y_continuous(  # noqa: F405
+        format="${,.0f}"
     )
-    + scale_fill_manual(values=colors)  # noqa: F405
     + labs(  # noqa: F405
-        x="Department", y="Salary ($)", title="violin-basic · lets-plot · pyplots.ai"
+        x="Department", y="Salary", title="violin-basic \u00b7 letsplot \u00b7 pyplots.ai"
     )
     + theme_minimal()  # noqa: F405
     + theme(  # noqa: F405
         axis_title=element_text(size=20),  # noqa: F405
         axis_text=element_text(size=16),  # noqa: F405
         plot_title=element_text(size=24),  # noqa: F405
-        legend_position="none",  # Legend not needed, x-axis shows categories
+        legend_position="none",
+        panel_grid_major_x=element_blank(),  # noqa: F405
+        axis_ticks=element_blank(),  # noqa: F405
     )
     + ggsize(1600, 900)  # noqa: F405
 )
 
-# Save PNG (scale 3x to get 4800 × 2700 px)
+# Save
 export_ggsave(plot, filename="plot.png", path=".", scale=3)
-
-# Save HTML for interactive version
 export_ggsave(plot, filename="plot.html", path=".")

plots/violin-basic/implementations/matplotlib.py

@@ -1,62 +1,81 @@
 """ pyplots.ai
 violin-basic: Basic Violin Plot
-Library: matplotlib 3.10.8 | Python 3.13.11
-Quality: 92/100 | Created: 2025-12-23
+Library: matplotlib 3.10.8 | Python 3.14.3
+Quality: 92/100 | Updated: 2026-02-21
 """
 
+import matplotlib.patheffects as pe
 import matplotlib.pyplot as plt
 import numpy as np
 
 
-# Data - simulated test scores across different schools
+# Data - simulated test scores (0-100) across different schools
 np.random.seed(42)
-categories = ["School A", "School B", "School C", "School D"]
+categories = ["Lincoln HS", "Roosevelt Acad.", "Jefferson HS", "Hamilton Prep"]
 data = [
-    np.random.normal(75, 10, 150),  # School A: centered around 75
-    np.random.normal(82, 8, 150),  # School B: higher scores, less spread
-    np.random.normal(68, 15, 150),  # School C: lower average, more spread
-    np.random.normal(78, 12, 150),  # School D: moderate
+    np.clip(np.random.normal(75, 10, 150), 0, 100),  # Lincoln: normal, centered ~75
+    np.clip(np.random.normal(85, 6, 150), 0, 100),  # Roosevelt: high, tight cluster
+    np.clip(np.random.normal(62, 15, 150), 0, 100),  # Jefferson: lower, wide spread
+    np.clip(
+        np.concatenate([np.random.normal(70, 5, 80), np.random.normal(88, 4, 70)]), 0, 100
+    ),  # Hamilton: bimodal (two subgroups)
 ]
 
-# Create plot (4800x2700 px)
+# Multi-series palette starting with Python Blue; warm accent for bimodal Hamilton
+colors = ["#306998", "#5BA58B", "#7A6FB5", "#D4853F"]
+edge_colors = ["#1E4060", "#3A7460", "#524A80", "#9A5F2A"]
+
+# Plot
 fig, ax = plt.subplots(figsize=(16, 9))
 
-# Create violin plot with quartile markers
-parts = ax.violinplot(data, positions=range(len(categories)), showmeans=False, showmedians=True, showextrema=True)
+parts = ax.violinplot(
+    data,
+    positions=range(len(categories)),
+    quantiles=[[0.25, 0.5, 0.75]] * len(categories),
+    showmeans=False,
+    showmedians=False,
+    showextrema=False,
+    bw_method=0.3,
+    widths=0.75,
+)
 
-# Style the violins with Python Blue
-for pc in parts["bodies"]:
-    pc.set_facecolor("#306998")
-    pc.set_edgecolor("#1e4a6e")
-    pc.set_alpha(0.7)
+# Style each violin body with a distinct color
+for i, pc in enumerate(parts["bodies"]):
+    pc.set_facecolor(colors[i])
+    pc.set_edgecolor(edge_colors[i])
+    pc.set_alpha(0.8)
     pc.set_linewidth(2)
 
-# Style the lines (median, min, max)
-parts["cmedians"].set_color("#FFD43B")
-parts["cmedians"].set_linewidth(3)
-parts["cmins"].set_color("#1e4a6e")
-parts["cmins"].set_linewidth(2)
-parts["cmaxes"].set_color("#1e4a6e")
-parts["cmaxes"].set_linewidth(2)
-parts["cbars"].set_color("#1e4a6e")
-parts["cbars"].set_linewidth(2)
-
-# Add quartile markers (Q1 and Q3) as box indicators
-quartile1 = [np.percentile(d, 25) for d in data]
-quartile3 = [np.percentile(d, 75) for d in data]
-
-# Draw thin boxes for interquartile range
-for i, (q1, q3) in enumerate(zip(quartile1, quartile3, strict=True)):
-    ax.vlines(i, q1, q3, color="#1e4a6e", linewidth=6, zorder=3)
-
-# Labels and styling (scaled font sizes for 4800x2700)
+# Quantile lines with path effects for legibility against colored bodies
+q_colors = ["white", "#FFD43B", "white"] * len(categories)
+q_widths = [2.5, 4, 2.5] * len(categories)
+parts["cquantiles"].set_colors(q_colors)
+parts["cquantiles"].set_linewidths(q_widths)
+parts["cquantiles"].set_path_effects([pe.Stroke(linewidth=6, foreground="black", alpha=0.3), pe.Normal()])
+
+# Labels and styling
 ax.set_xticks(range(len(categories)))
 ax.set_xticklabels(categories)
 ax.set_xlabel("School", fontsize=20)
 ax.set_ylabel("Test Score (points)", fontsize=20)
-ax.set_title("violin-basic · matplotlib · pyplots.ai", fontsize=24)
+ax.set_title("violin-basic \u00b7 matplotlib \u00b7 pyplots.ai", fontsize=24, fontweight="medium")
 ax.tick_params(axis="both", labelsize=16)
-ax.grid(True, alpha=0.3, linestyle="--", axis="y")
+
+ax.spines["top"].set_visible(False)
+ax.spines["right"].set_visible(False)
+ax.yaxis.grid(True, alpha=0.2, linewidth=0.8)
+
+# Subtle annotation highlighting Hamilton Prep's bimodal distribution
+ax.annotate(
+    "Two distinct\nperformance groups",
+    xy=(3, 75),
+    xytext=(3, 45),
+    fontsize=13,
+    color="#9A5F2A",
+    fontstyle="italic",
+    ha="center",
+    arrowprops={"arrowstyle": "->", "color": "#9A5F2A", "lw": 1.5},
+)
 
 plt.tight_layout()
 plt.savefig("plot.png", dpi=300, bbox_inches="tight")