1"""Phase 7 — publication figures (vector PDF) → paper/figures/. Palette and mark rules follow the dataviz reference2(categorical slots in fixed order, one hue sequential ramp, thin marks, direct labels, recessive grid, no dual axes)."""3import json, numpy as np, pandas as pd, matplotlib4matplotlib.use("Agg")5import matplotlib.pyplot as plt6from matplotlib.colors import LinearSegmentedColormap7from src.common import ROOT, path, get_logger8log = get_logger("figures"); P = path("processed"); F = ROOT / "paper" / "figures"; E = ROOT / "experiments"; F.mkdir(exist_ok=True)9C = ["#2a78d6", "#eb6834", "#1baf7a", "#eda100", "#e87ba4", "#008300", "#6250d6", "#e34948"]10SEQ = LinearSegmentedColormap.from_list("seq", ["#cde2fb", "#86b6ef", "#3987e5", "#256abf", "#184f95", "#0d366b"])11TXT, TXT2, GRID = "#0b0b0b", "#52514e", "#e6e5e1"12plt.rcParams.update({"font.size": 9, "axes.edgecolor": TXT2, "axes.labelcolor": TXT, "xtick.color": TXT2, "ytick.color": TXT2, "axes.spines.top": False, "axes.spines.right": False,13 "axes.grid": True, "grid.color": GRID, "grid.linewidth": 0.6, "axes.axisbelow": True, "legend.frameon": False, "pdf.fonttype": 42})1415def save(fig, name):16 fig.savefig(F / f"{name}.pdf", bbox_inches="tight"); plt.close(fig); log.info("wrote %s.pdf", name)1718def fig_coverage():19 wk = pd.read_csv(P / "coverage_by_week.csv", parse_dates=["week"]); feed = pd.read_csv(P / "coverage_by_feed.csv"); feed = feed[feed.active > 0]20 fig, axes = plt.subplots(1, 2, figsize=(7.2, 3.0), gridspec_kw={"width_ratios": [1.3, 1]})21 ax = axes[0]; ax.plot(wk.week, wk.online, color=C[0], lw=2, marker="o", ms=4); ax.plot(wk.week, wk.circular, color=C[1], lw=2, marker="o", ms=4)22 ax.text(wk.week.iloc[-1], wk.online.iloc[-1], " storefront", color=TXT, va="center"); ax.text(wk.week.iloc[-1], wk.circular.iloc[-1], " circular", color=TXT, va="center")23 ax.set_ylabel("Listings observed in week"); ax.set_title("A. Weekly coverage by channel", loc="left", color=TXT); ax.tick_params(axis="x", rotation=30); ax.set_ylim(0)24 ax = axes[1]; comp = feed.groupby("company").active.sum().sort_values(ascending=True).tail(10)25 ax.barh(comp.index, comp.values, color=C[0], height=0.6); ax.set_title("B. Active listings by company (top 10)", loc="left", color=TXT); ax.set_xlabel("Active listings, 1 Nov 2026")26 for i, v in enumerate(comp.values): ax.text(v, i, f" {v:,}", va="center", color=TXT2, fontsize=8)27 ax.grid(axis="y", visible=False)28 save(fig, "fig_coverage")2930def fig_attrition():31 att = pd.read_csv(P / "attrition_listings.csv"); v = pd.read_csv(P / "match_validation_summary.csv")32 fig, axes = plt.subplots(1, 2, figsize=(7.2, 3.0), gridspec_kw={"width_ratios": [1.4, 1]})33 ax = axes[0]; y = np.arange(len(att))[::-1]; ax.barh(y, att.listings, color=[C[0]] * (len(att) - 1) + [C[2]], height=0.6)34 labels = ["All listings", "excl. SAQ", "excl. pharmacies", "excl. non-food", "with a price", "excl. unrealistic", "parsable size", "matched (A–C)"]35 ax.set_yticks(y); ax.set_yticklabels(labels); ax.set_xlabel("Listings"); ax.set_title("A. Attrition flow", loc="left", color=TXT); ax.grid(axis="y", visible=False)36 for yi, n in zip(y, att.listings): ax.text(n, yi, f" {n:,}", va="center", color=TXT2, fontsize=8)37 ax = axes[1]; x = np.arange(4); ax.bar(x - 0.2, v.precision_strict, 0.38, color=C[0], label="Text review (240 pairs)"); ax.bar(x + 0.2, v.upc_precision.fillna(0), 0.38, color=C[1], label="UPC ground truth")38 ax.errorbar(x - 0.2, v.precision_strict, yerr=[v.precision_strict - v.ci_lo, v.ci_hi - v.precision_strict], fmt="none", ecolor=TXT2, lw=1)39 ax.set_xticks(x); ax.set_xticklabels([f"Tier {t}" for t in v.tier]); ax.set_ylim(0, 1.05); ax.set_ylabel("Precision"); ax.set_title("B. Matching precision by tier", loc="left", color=TXT); ax.legend(loc="lower left", fontsize=7); ax.grid(axis="x", visible=False)40 save(fig, "fig_matching")4142def fig_gaps():43 st = pd.read_csv(P / "modelA_groupweek_stats.csv"); fig, axes = plt.subplots(1, 2, figsize=(7.2, 3.0))44 ax = axes[0]; bins = np.linspace(0, 1.0, 41)45 ax.hist(st[st.same_company].gap.clip(upper=1), bins=bins, color=C[0], alpha=0.85, density=True, label="same company"); ax.hist(st[~st.same_company].gap.clip(upper=1), bins=bins, color=C[1], alpha=0.6, density=True, label="different companies")46 ax.set_xlabel("Within-product max–min log price gap (group-week)"); ax.set_ylabel("Density"); ax.legend(); ax.set_title("A. Distribution of cross-banner gaps", loc="left", color=TXT); ax.grid(axis="x", visible=False)47 fe = pd.read_csv(P / "modelA_banner_effects.csv").sort_values("coef"); fe = fe[fe.n_listing_weeks >= 50]48 ax = axes[1]; y = np.arange(len(fe)); ax.errorbar(fe.coef, y, xerr=1.96 * fe.se, fmt="o", color=C[0], ecolor=TXT2, ms=4, lw=1); ax.axvline(0, color=TXT2, lw=0.8)49 ax.set_yticks(y); ax.set_yticklabels(fe.banner); ax.set_xlabel("Banner effect on log price (ref. Metro), 95% CI"); ax.set_title("B. Relative price levels", loc="left", color=TXT); ax.grid(axis="y", visible=False)50 save(fig, "fig_dispersion")5152def fig_heatmap():53 pt = pd.read_csv(P / "modelA_banner_pairs.csv"); pt = pt[pt.n >= 30]54 banners = sorted(set(pt.banner_r) | set(pt.banner_s)); M = pd.DataFrame(np.nan, index=banners, columns=banners)55 for _, r in pt.iterrows(): M.at[r.banner_r, r.banner_s] = r.median_gap; M.at[r.banner_s, r.banner_r] = -r.median_gap56 fig, ax = plt.subplots(figsize=(5.6, 4.8)); cm = LinearSegmentedColormap.from_list("div", ["#2a78d6", "#f0efec", "#e34948"])57 im = ax.imshow(M.values, cmap=cm, vmin=-0.3, vmax=0.3); ax.set_xticks(range(len(banners))); ax.set_xticklabels(banners, rotation=60, ha="right"); ax.set_yticks(range(len(banners))); ax.set_yticklabels(banners); ax.grid(False)58 for i in range(len(banners)):59 for j in range(len(banners)):60 if not np.isnan(M.values[i, j]): ax.text(j, i, f"{M.values[i, j]:+.2f}", ha="center", va="center", fontsize=6.5, color=TXT)61 cb = fig.colorbar(im, ax=ax, fraction=0.04); cb.set_label("Median log price gap, row banner minus column banner")62 ax.set_title("Pairwise median gaps on identical products (pairs with ≥ 30 group-weeks)", loc="left", color=TXT, fontsize=9)63 save(fig, "fig_heatmap")6465def fig_promotions():66 es = pd.read_csv(P / "modelB_event_study.csv"); es["k"] = es.term.str.extract(r"(-?\d+)$").astype(int); es = pd.concat([es, pd.DataFrame([{"term": "ref", "coef": 0, "se": 0, "k": -1}])]).sort_values("k")67 ep = pd.read_parquet(P / "modelB_episodes.parquet"); tm = pd.read_csv(P / "modelB_transition_matrix.csv", index_col=0)68 fig, axes = plt.subplots(1, 3, figsize=(7.6, 2.8), gridspec_kw={"width_ratios": [1.3, 1, 0.8]})69 ax = axes[0]; ax.fill_between(es.k, es.coef - 1.96 * es.se, es.coef + 1.96 * es.se, color=C[0], alpha=0.18, lw=0); ax.plot(es.k, es.coef, color=C[0], lw=2, marker="o", ms=4); ax.axhline(0, color=TXT2, lw=0.8); ax.axvline(-0.5, color=TXT2, lw=0.8, ls=":")70 ax.set_xlabel("Weeks relative to first promoted week"); ax.set_ylabel("Log price vs week −1"); ax.set_title("A. Price path around temporary sales", loc="left", color=TXT)71 ax = axes[1]; unc = ep[~ep.censored]; ax.hist(unc[unc.kind == "return"].depth, bins=np.linspace(0, 1.2, 31), color=C[0], alpha=0.85, label="return (temporary)"); ax.hist(unc[unc.kind == "reset_down"].depth, bins=np.linspace(0, 1.2, 31), color=C[1], alpha=0.6, label="reset down")72 ax.set_xlabel("Episode depth, log(reference / price)"); ax.set_ylabel("Episodes"); ax.legend(fontsize=7); ax.set_title("B. Depth by episode kind", loc="left", color=TXT); ax.grid(axis="x", visible=False)73 ax = axes[2]; im = ax.imshow(tm.values, cmap=SEQ, vmin=0, vmax=1); ax.set_xticks([0, 1]); ax.set_xticklabels(["regular", "promo"]); ax.set_yticks([0, 1]); ax.set_yticklabels(["regular", "promo"]); ax.set_xlabel("state in week w"); ax.set_ylabel("state in week w−1"); ax.grid(False)74 for i in range(2):75 for j in range(2): ax.text(j, i, f"{tm.values[i, j]:.2f}", ha="center", va="center", color="white" if tm.values[i, j] > 0.5 else TXT)76 ax.set_title("C. Weekly transitions", loc="left", color=TXT)77 save(fig, "fig_promotions")7879def fig_spillovers():80 d = pd.read_csv(P / "modelC_distributed_lags.csv"); fig, ax = plt.subplots(figsize=(5.2, 2.8))81 order = {"rival_all": 0, "rival_same": 1, "rival_cross": 2}; lab = {"rival_all": "all rivals", "rival_same": "same company", "rival_cross": "other company"}82 xs = ["_f2", "_f1", "", "_l1", "_l2", "_l3"]; xl = ["lead 2", "lead 1", "k=0", "lag 1", "lag 2", "lag 3"]83 for col, g in d.groupby("rival"):84 i = order[col]; vals = [g[g.term == col + s].coef.iloc[0] for s in xs]; ses = [g[g.term == col + s].se.iloc[0] for s in xs]85 ax.errorbar(np.arange(6) + (i - 1) * 0.22, vals, yerr=[1.96 * v for v in ses], fmt="o", color=C[i], ms=4, lw=1, label=lab[col])86 ax.axhline(0, color=TXT2, lw=0.8); ax.axvspan(-0.5, 1.5, color=GRID, alpha=0.5, lw=0); ax.set_xticks(range(6)); ax.set_xticklabels(xl); ax.set_ylabel("Response of own Δlog p"); ax.legend(fontsize=7, ncol=3, loc="upper left")87 ax.set_title("Lead–lag association with rivals' price changes (shaded = placebo leads)", loc="left", color=TXT, fontsize=9); ax.grid(axis="x", visible=False)88 save(fig, "fig_spillovers")8990def fig_indices():91 d = pd.read_csv(P / "modelD_indices.csv", parse_dates=["week"]); s = d[d["sample"] == "storefront_posted"]92 fig, ax = plt.subplots(figsize=(5.6, 3.0))93 ax.plot(s.week, s.index_mean, color=C[1], lw=2, marker="o", ms=3); ax.plot(s.week, s["index"], color=C[0], lw=2, marker="o", ms=3); ax.plot(s.week, s.jevons, color=C[2], lw=2, marker="o", ms=3, ls="--"); ax.plot(s.week, s.carli, color=C[3], lw=1.5, marker="o", ms=3, ls=":")94 for col, name in [("index_mean", "naive mean of posted prices"), ("index", "chained Jevons (matched model)"), ("jevons", "fixed-basket Jevons"), ("carli", "fixed-basket Carli")]:95 ax.text(s.week.iloc[-1], s[col].iloc[-1], " " + name, va="center", fontsize=7.5, color=TXT)96 ax.set_ylabel("Index (13 Aug 2026 = 100)"); ax.tick_params(axis="x", rotation=30); ax.set_title("Unweighted price indices, storefront listings", loc="left", color=TXT, fontsize=9)97 save(fig, "fig_indices")9899def fig_qgpai():100 q = pd.read_csv(P / "qgpai_components.csv", parse_dates=["week"]); fig, axes = plt.subplots(2, 2, figsize=(7.2, 4.8)); axes = axes.ravel()101 ax = axes[0]; ax.fill_between(q.week, q.cpi_lo, q.cpi_hi, color=C[0], alpha=0.18, lw=0); ax.plot(q.week, q.comparable_price_index, color=C[0], lw=2, marker="o", ms=3); ax.set_title("1. Comparable price index (matched groups)", loc="left", color=TXT); ax.set_ylabel("13 Aug = 100")102 ax = axes[1]; ax.plot(q.week, q.promo_share, color=C[1], lw=2, marker="o", ms=3); ax.plot(q.week, q.circular_depth_median, color=C[3], lw=2, marker="s", ms=3)103 ax.text(q.week.iloc[-1], q.promo_share.iloc[-1], " share promoted\n (storefront, inferred)", fontsize=7, va="center", color=TXT); ax.text(q.week.iloc[-1], q.circular_depth_median.iloc[-1], " circular depth\n (observed, median log)", fontsize=7, va="center", color=TXT)104 ax.set_title("2. Promotion intensity", loc="left", color=TXT); ax.set_ylim(0)105 ax = axes[2]; ax.fill_between(q.week, q.disp_lo, q.disp_hi, color=C[2], alpha=0.18, lw=0); ax.plot(q.week, q.dispersion_median_gap, color=C[2], lw=2, marker="o", ms=3); ax.set_title("3. Retail price dispersion (median log gap)", loc="left", color=TXT); ax.set_ylim(0)106 ax = axes[3]107 for col, name, c in [("basket_K1", "best single banner", C[0]), ("basket_K2", "two banners", C[1]), ("basket_unconstrained", "unconstrained", C[2])]:108 ax.plot(q.week, q[col], color=c, lw=2, marker="o", ms=3); ax.text(q.week.iloc[-1], q[col].iloc[-1], " " + name, fontsize=7, va="center", color=TXT)109 ax.set_title("4. Feasible basket cost (CAD, fixed block)", loc="left", color=TXT); ax.set_ylim(0)110 for ax in axes: ax.tick_params(axis="x", rotation=30)111 fig.suptitle("QGPAI components (component 5, access & affordability, not computable)", x=0.01, ha="left", color=TXT2, fontsize=9)112 save(fig, "fig_qgpai")113114def fig_basket():115 r = pd.read_csv(P / "modelE_results.csv"); r5 = r[(r.B == 5) & r.single.isna() & r.fixed_cost.isna()]; fr = r5.groupby("K").cost.mean(); ks = [1, 2, 3, 4, 5]; costs = [fr.get(k, fr.get(99)) if k < 5 else fr.get(99) for k in ks]116 tc = pd.read_csv(P / "modelE_travel_sensitivity.csv")117 fig, axes = plt.subplots(1, 2, figsize=(7.2, 2.8))118 ax = axes[0]; ax.plot(ks, costs, color=C[0], lw=2, marker="o", ms=5); ax.set_xlabel("Maximum number of banners visited (K)"); ax.set_ylabel("Weekly basket cost (CAD, mean of 11 weeks)"); ax.set_xticks(ks)119 for k, c in zip(ks, costs): ax.text(k, c, f" {c:.1f}", fontsize=7.5, color=TXT2, va="bottom")120 ax.set_title("A. Savings frontier (B = 5, 15 items)", loc="left", color=TXT)121 ax = axes[1]; ax.bar(tc.fixed_cost.astype(str), tc.cost_incl, color=C[1], width=0.55); ax.bar(tc.fixed_cost.astype(str), tc.cost_goods, color=C[0], width=0.55)122 for i, (g, t, n) in enumerate(zip(tc.cost_goods, tc.cost_incl, tc.n_used)): ax.text(i, t, f" {n:.1f} banners", ha="center", va="bottom", fontsize=7.5, color=TXT2)123 ax.set_xlabel("Illustrative fixed cost per banner visited (CAD)"); ax.set_ylabel("Goods (blue) + travel (orange)"); ax.set_title("B. Travel-cost sensitivity (assumption-based)", loc="left", color=TXT); ax.grid(axis="x", visible=False)124 save(fig, "fig_basket")125126def fig_forecast():127 r = pd.read_csv(P / "modelF_rolling_results.csv"); fig, axes = plt.subplots(1, 2, figsize=(7.2, 2.8))128 ax = axes[0]; d = r[r.target == "log_price"]129 for i, (m, g) in enumerate(d.groupby("model")): ax.plot(g.origin_t, g.mae, color=C[i % 8], lw=1.8, marker="o", ms=3, label=m)130 ax.set_xlabel("Forecast origin (week index)"); ax.set_ylabel("MAE of next-week log price"); ax.legend(fontsize=6.5, ncol=2); ax.set_title("A. Price forecasts", loc="left", color=TXT)131 ax = axes[1]; d = r[r.target == "promotion"]132 for i, (m, g) in enumerate(d.groupby("model")): ax.plot(g.origin_t, g.auc, color=C[i % 8], lw=1.8, marker="o", ms=3, label=m)133 ax.set_xlabel("Forecast origin (week index)"); ax.set_ylabel("AUC, next-week promotion"); ax.set_ylim(0.4, 1); ax.legend(fontsize=6.5, ncol=2, loc="lower right"); ax.set_title("B. Promotion forecasts", loc="left", color=TXT)134 save(fig, "fig_forecast")135136def fig_store_mixture():137 d = pd.read_parquet(P / "offers_daily.parquet", columns=["uid", "day", "p_min", "p_max", "p_mode", "n_distinct"]); x = d[d.uid == "metro:068400662600"].sort_values("day")138 pl = pd.read_parquet(path("interim") / "raw_price_log.parquet"); e = pl[pl.uid == "metro:068400662600"].copy(); e["ts"] = pd.to_datetime(e.ts, unit="s", utc=True).dt.tz_convert("America/Toronto")139 sm = pd.read_csv(P / "modelA_within_banner_mixture.csv").sort_values("share_multi", ascending=True).tail(12)140 fig, axes = plt.subplots(1, 2, figsize=(7.4, 2.9), gridspec_kw={"width_ratios": [1.4, 1]})141 ax = axes[0]; ax.step(e.ts, e.price, where="post", color=C[0], lw=1.2); ax.scatter(e.ts, e.price, color=C[0], s=8, zorder=3)142 ax.set_ylabel("Posted price (CAD)"); ax.set_title("A. One listing (Metro, mayonnaise 890 mL): every logged price change", loc="left", color=TXT, fontsize=8.5); ax.tick_params(axis="x", rotation=30)143 ax = axes[1]; ax.barh(sm.banner, sm.share_multi, color=C[0], height=0.6); ax.set_xlabel("Share of listing-weeks with > 1 price value"); ax.set_title("B. Multi-price weeks by banner", loc="left", color=TXT); ax.grid(axis="y", visible=False)144 for i, (v, r_) in enumerate(zip(sm.share_multi, sm.median_log_range_multi)): ax.text(v, i, f" {v:.2f} (range {r_:.2f})", va="center", fontsize=7, color=TXT2)145 save(fig, "fig_store_mixture")146147def main():148 for f in [fig_coverage, fig_attrition, fig_gaps, fig_heatmap, fig_promotions, fig_spillovers, fig_indices, fig_qgpai, fig_basket, fig_forecast, fig_store_mixture]:149 try: f()150 except Exception as ex: log.error("%s failed: %s", f.__name__, ex)151152if __name__ == "__main__":153 main()154