Ajoute la richesse chromatique par set

lib/plots/color_richness.py

@@ -0,0 +1,130 @@
+"""Visualisations de la richesse chromatique par set."""
+
+from pathlib import Path
+from typing import Iterable, List, Tuple
+
+import matplotlib.pyplot as plt
+import numpy as np
+from matplotlib.patches import Patch
+
+from lib.filesystem import ensure_parent_dir
+from lib.rebrickable.stats import read_rows
+
+
+def load_richness_rows(path: Path) -> List[dict]:
+    """Charge les métriques de richesse chromatique."""
+    return read_rows(path)
+
+
+def build_boxplot_data(rows: Iterable[dict]) -> Tuple[List[List[int]], List[str]]:
+    """Prépare les valeurs de boxplot par année."""
+    grouped: dict[str, List[int]] = {}
+    for row in rows:
+        year_rows = grouped.get(row["year"])
+        if year_rows is None:
+            year_rows = []
+            grouped[row["year"]] = year_rows
+        year_rows.append(int(row["colors_distinct"]))
+    years = sorted(grouped.keys(), key=int)
+    data = [grouped[year] for year in years]
+    return data, years
+
+
+def plot_richness_boxplot(richness_path: Path, destination_path: Path) -> None:
+    """Trace le boxplot du nombre de couleurs distinctes par set et par année."""
+    rows = load_richness_rows(richness_path)
+    if not rows:
+        return
+    data, years = build_boxplot_data(rows)
+    fig, ax = plt.subplots(figsize=(12, 7))
+    box = ax.boxplot(
+        data,
+        orientation="vertical",
+        patch_artist=True,
+        tick_labels=years,
+        boxprops=dict(facecolor="#1f77b4", alpha=0.3),
+        medianprops=dict(color="#0d0d0d", linewidth=1.5),
+        whiskerprops=dict(color="#555555", linestyle="--"),
+        capprops=dict(color="#555555"),
+    )
+    for patch in box["boxes"]:
+        patch.set_edgecolor("#1f77b4")
+    ax.set_xlabel("Année")
+    ax.set_ylabel("Nombre de couleurs distinctes (hors rechanges)")
+    ax.set_title("Richesse chromatique par set (répartition annuelle)")
+    ax.grid(axis="y", linestyle="--", alpha=0.3)
+
+    ensure_parent_dir(destination_path)
+    fig.tight_layout()
+    fig.savefig(destination_path, dpi=170)
+    plt.close(fig)
+
+
+def select_top_sets(rows: Iterable[dict], limit: int = 15) -> List[dict]:
+    """Retient les sets les plus colorés et les plus concentrés."""
+    sorted_rows = sorted(
+        rows,
+        key=lambda row: (-int(row["colors_distinct"]), float(row["top3_share"]), row["set_num"]),
+    )
+    return sorted_rows[:limit]
+
+
+def plot_richness_top_sets(richness_path: Path, destination_path: Path) -> None:
+    """Trace le top des sets les plus riches en couleurs."""
+    rows = load_richness_rows(richness_path)
+    if not rows:
+        return
+    top_rows = select_top_sets(rows)
+    y_positions = np.arange(len(top_rows))
+    counts = [int(row["colors_distinct"]) for row in top_rows]
+    labels = [f"{row['set_num']} · {row['name']} ({row['year']})" for row in top_rows]
+    owned_mask = [row["in_collection"] == "true" for row in top_rows]
+
+    fig, ax = plt.subplots(figsize=(11, 8))
+    for y, value, owned in zip(y_positions, counts, owned_mask):
+        alpha = 0.92 if owned else 0.45
+        ax.barh(y, value, color="#2ca02c", alpha=alpha)
+    ax.set_yticks(y_positions)
+    ax.set_yticklabels(labels)
+    ax.invert_yaxis()
+    ax.set_xlabel("Couleurs distinctes (hors rechanges)")
+    ax.set_title("Top des sets les plus colorés")
+    ax.grid(axis="x", linestyle="--", alpha=0.3)
+    legend = [
+        Patch(facecolor="#2ca02c", edgecolor="none", alpha=0.92, label="Set possédé"),
+        Patch(facecolor="#2ca02c", edgecolor="none", alpha=0.45, label="Set manquant"),
+    ]
+    ax.legend(handles=legend, loc="lower right", frameon=False)
+
+    ensure_parent_dir(destination_path)
+    fig.tight_layout()
+    fig.savefig(destination_path, dpi=170)
+    plt.close(fig)
+
+
+def plot_concentration_scatter(richness_path: Path, destination_path: Path) -> None:
+    """Visualise la concentration de palette vs nombre de couleurs."""
+    rows = load_richness_rows(richness_path)
+    if not rows:
+        return
+    x_values = [int(row["colors_distinct"]) for row in rows]
+    y_values = [float(row["top3_share"]) for row in rows]
+    owned_mask = [row["in_collection"] == "true" for row in rows]
+    colors = ["#1f77b4" if owned else "#bbbbbb" for owned in owned_mask]
+
+    fig, ax = plt.subplots(figsize=(10, 7))
+    ax.scatter(x_values, y_values, c=colors, alpha=0.7, s=32)
+    ax.set_xlabel("Nombre de couleurs distinctes (hors rechanges)")
+    ax.set_ylabel("Part des 3 couleurs principales")
+    ax.set_title("Concentration des palettes")
+    ax.grid(True, linestyle="--", alpha=0.3)
+    legend = [
+        Patch(facecolor="#1f77b4", edgecolor="none", alpha=0.7, label="Set possédé"),
+        Patch(facecolor="#bbbbbb", edgecolor="none", alpha=0.7, label="Set manquant"),
+    ]
+    ax.legend(handles=legend, loc="upper right", frameon=False)
+
+    ensure_parent_dir(destination_path)
+    fig.tight_layout()
+    fig.savefig(destination_path, dpi=170)
+    plt.close(fig)