"""One-shot analysis + visualisation of the CNIL sanctions dataset. Produces: - charts/yearly_volume.png - charts/yearly_fines.png - charts/sector_breakdown.png - charts/breach_themes.png - charts/fine_buckets.png - charts/simplified_share.png - insights.json (machine-readable headline numbers for the README) """ from __future__ import annotations import json from collections import Counter from pathlib import Path import matplotlib matplotlib.use("Agg") import matplotlib.pyplot as plt import pandas as pd HERE = Path(__file__).parent PARQUET = HERE / "cnil_sanctions_analysis.parquet" CHARTS = HERE / "charts" CHARTS.mkdir(exist_ok=True) # ── Palette + style ───────────────────────────────────────────────────────── plt.rcParams.update({ "figure.dpi": 130, "savefig.dpi": 130, "savefig.bbox": "tight", "font.family": "DejaVu Sans", "axes.spines.top": False, "axes.spines.right": False, "axes.grid": True, "axes.grid.axis": "y", "grid.color": "#e5e7eb", "grid.linestyle": "-", "grid.linewidth": 0.8, "axes.titlesize": 14, "axes.titleweight": "bold", "axes.labelsize": 11, }) ACCENT = "#2563eb" ACCENT_2 = "#dc2626" NEUTRAL = "#6b7280" df = pd.read_parquet(PARQUET) print(f"loaded {len(df)} rows, {len(df.columns)} cols") # ── 1. Yearly volume ──────────────────────────────────────────────────────── yearly = df["n_sanction_year"].dropna().astype(int).value_counts().sort_index() yearly_simplified = ( df[df["is_simplified_procedure"] == True]["n_sanction_year"] .dropna() .astype(int) .value_counts() .reindex(yearly.index, fill_value=0) ) yearly_standard = yearly - yearly_simplified fig, ax = plt.subplots(figsize=(10, 5)) ax.bar(yearly.index, yearly_standard, color=ACCENT, label="standard / unspecified procedure") ax.bar(yearly.index, yearly_simplified, bottom=yearly_standard, color=ACCENT_2, label="simplified procedure") ax.set_title("CNIL sanctions per year, by procedure type (2011 – 2025)") ax.set_xlabel("Year") ax.set_ylabel("Number of sanctions") ax.set_xticks(yearly.index) ax.legend(frameon=False, loc="upper left") for x, y in zip(yearly.index, yearly.values): ax.text(x, y + 1, str(int(y)), ha="center", va="bottom", fontsize=8, color=NEUTRAL) fig.savefig(CHARTS / "yearly_volume.png") plt.close(fig) print(" ✓ yearly_volume.png") # ── 2. Yearly fine totals + sanction counts ──────────────────────────────── df_fines = df.dropna(subset=["amount_fine_eur", "n_sanction_year"]).copy() df_fines["n_sanction_year"] = df_fines["n_sanction_year"].astype(int) fines_per_year = df_fines.groupby("n_sanction_year")["amount_fine_eur"].sum() / 1e6 fines_per_year = fines_per_year.reindex(yearly.index, fill_value=0.0) fines_count = df_fines.groupby("n_sanction_year").size().reindex(yearly.index, fill_value=0) fig, ax = plt.subplots(figsize=(10, 5)) ax.bar(fines_per_year.index, fines_per_year.values, color=ACCENT) ax.set_title("Aggregate disclosed fines per year (€M)") ax.set_xlabel("Year") ax.set_ylabel("Fines (€ millions)") ax.set_xticks(fines_per_year.index) for x, y, c in zip(fines_per_year.index, fines_per_year.values, fines_count.values): if y > 0: ax.text(x, y + 5, f"€{y:.0f}M\n(n={c})", ha="center", va="bottom", fontsize=8, color=NEUTRAL) fig.savefig(CHARTS / "yearly_fines.png") plt.close(fig) print(" ✓ yearly_fines.png") # ── 3. Sector breakdown ──────────────────────────────────────────────────── sec = df["cat_sector_group"].dropna().value_counts() labels_map = { "private_company": "Private company", "public": "Public sector", "professional_individual": "Professional individual", "association": "Association", "political": "Political party", "other": "Other", } sec = sec.rename(index=lambda x: labels_map.get(x, x)) fig, ax = plt.subplots(figsize=(8, 5)) bars = ax.barh(sec.index[::-1], sec.values[::-1], color=ACCENT) ax.set_title("Sanctions by sector (2011 – 2025)") ax.set_xlabel("Number of sanctions") ax.grid(axis="y", visible=False) ax.grid(axis="x", visible=True) for bar, val in zip(bars, sec.values[::-1]): ax.text(val + 3, bar.get_y() + bar.get_height() / 2, f"{val} ({val / sec.sum():.0%})", va="center", fontsize=9, color=NEUTRAL) fig.savefig(CHARTS / "sector_breakdown.png") plt.close(fig) print(" ✓ sector_breakdown.png") # ── 4. Breach themes ─────────────────────────────────────────────────────── breach_cols = [c for c in df.columns if c.startswith("is_breach_") or c == "is_involves_sensitive_data"] breach_counts = {c: int(df[c].fillna(False).sum()) for c in breach_cols} breach_counts = dict(sorted(breach_counts.items(), key=lambda kv: kv[1])) pretty = { "is_breach_security": "Security of processing (Art. 32)", "is_breach_transparency": "Information / transparency", "is_breach_consent": "Lawful basis: consent", "is_breach_data_rights": "Rights of data subjects", "is_breach_cookies": "Cookies / trackers (Art. 82)", "is_breach_minimization": "Data minimisation", "is_breach_processor": "Sub-processor obligations", "is_involves_sensitive_data": "Special-category data (Art. 9)", } labels = [pretty.get(c, c) for c in breach_counts] vals = list(breach_counts.values()) fig, ax = plt.subplots(figsize=(9, 5)) bars = ax.barh(labels, vals, color=ACCENT) ax.set_title("Breach themes detected across CNIL sanctions") ax.set_xlabel(f"# of decisions mentioning this theme (n={len(df)})") ax.grid(axis="y", visible=False) ax.grid(axis="x", visible=True) for bar, v in zip(bars, vals): ax.text(v + 2, bar.get_y() + bar.get_height() / 2, f"{v} ({v / len(df):.0%})", va="center", fontsize=9, color=NEUTRAL) fig.savefig(CHARTS / "breach_themes.png") plt.close(fig) print(" ✓ breach_themes.png") # ── 5. Fine bucket distribution ──────────────────────────────────────────── fb = df["cat_fine_bucket"].dropna().value_counts() order = ["none", "under_10k", "under_100k", "under_1m", "over_1m"] fb = fb.reindex([o for o in order if o in fb.index], fill_value=0) label_order = { "none": "no fine", "under_10k": "< €10 k", "under_100k": "€10 k – 100 k", "under_1m": "€100 k – 1 M", "over_1m": "> €1 M", } fb = fb.rename(index=label_order) fig, ax = plt.subplots(figsize=(8, 5)) ax.bar(fb.index, fb.values, color=ACCENT) ax.set_title("Fine-size distribution") ax.set_xlabel("Fine bucket") ax.set_ylabel("Number of sanctions") for i, (idx, v) in enumerate(fb.items()): ax.text(i, v + 1, str(int(v)), ha="center", va="bottom", fontsize=9, color=NEUTRAL) fig.savefig(CHARTS / "fine_buckets.png") plt.close(fig) print(" ✓ fine_buckets.png") # ── 6. Simplified-procedure share over time ──────────────────────────────── simp_share = (yearly_simplified / yearly.replace(0, 1)).fillna(0) * 100 fig, ax = plt.subplots(figsize=(10, 4.5)) ax.plot(simp_share.index, simp_share.values, color=ACCENT_2, marker="o", linewidth=2) ax.fill_between(simp_share.index, 0, simp_share.values, color=ACCENT_2, alpha=0.15) ax.set_title("Share of sanctions issued under the SIMPLIFIED procedure") ax.set_xlabel("Year") ax.set_ylabel("% of yearly sanctions") ax.set_xticks(simp_share.index) ax.set_ylim(0, 100) for x, y in zip(simp_share.index, simp_share.values): if y > 0: ax.text(x, y + 2, f"{y:.0f}%", ha="center", fontsize=8, color=NEUTRAL) fig.savefig(CHARTS / "simplified_share.png") plt.close(fig) print(" ✓ simplified_share.png") # ── insights ─────────────────────────────────────────────────────────────── fines_clean = df["amount_fine_eur"].dropna() top_fines = ( df.dropna(subset=["amount_fine_eur"]) .sort_values("amount_fine_eur", ascending=False) .head(10)[["dn_sanction", "organism_type_raw", "amount_fine_eur", "main_breaches_raw"]] .to_dict(orient="records") ) insights = { "n_rows": int(len(df)), "years": [int(df["n_sanction_year"].min()), int(df["n_sanction_year"].max())], "growth_2014_to_2024": round(yearly.get(2024, 0) / max(1, yearly.get(2014, 1)), 1), "growth_2022_to_2024": round(yearly.get(2024, 0) / max(1, yearly.get(2022, 1)), 1), "private_share_pct": round( 100 * (df["cat_sector_group"] == "private_company").sum() / len(df), 1 ), "simplified_share_total_pct": round( 100 * (df["is_simplified_procedure"] == True).sum() / max(1, df["is_simplified_procedure"].notna().sum()), 1, ), "simplified_share_2024_pct": round(simp_share.get(2024, 0.0), 1), "fines_reported": int(fines_clean.shape[0]), "total_fines_meur": round(fines_clean.sum() / 1e6, 1), "median_fine_eur": float(fines_clean.median()) if len(fines_clean) else 0.0, "max_fine_eur": float(fines_clean.max()) if len(fines_clean) else 0.0, "top_breach_theme": max(breach_counts, key=breach_counts.get), "top_breach_count": max(breach_counts.values()), "top_fines": top_fines, } (HERE / "insights.json").write_text( json.dumps(insights, indent=2, ensure_ascii=False, default=str) ) print("\ninsights:", json.dumps(insights, indent=2, ensure_ascii=False, default=str))