Upload subgroup shared-basis stability analysis

evaluation/query_family/subgroup/shared_basis_stability_analysis/INSIGHTS.md

@@ -0,0 +1,53 @@
+# Shared-Basis Stability Insights
+
+## Main reading
+
+On the shared support-oriented basis `support3_score = mean(key_set_score, row_count_score, column_score)`, subgroup-size queries are more robust than internal-profile queries.
+
+- Query-level means: `internal = 0.777`, `size = 0.818`
+- Query-level medians: `internal = 0.889`, `size = 1.000`
+- Low-tail share `<= 0.4`: `internal = 0.143`, `size = 0.129`
+- High-score share `>= 0.8`: `internal = 0.538`, `size = 0.632`
+
+## Matched-panel interpretation
+
+Comparing the two subgroup branches inside the same dataset-model panel avoids raw-score incomparability. Under this matched view, `support3(size) - support3(internal)` is positive on average.
+
+- Mean matched-panel delta: `0.011`
+- Share of panels with `size > internal`: `0.437`
+- Share of panels with `size - internal >= 0.10`: `0.126`
+- This is therefore a distribution-level stability story, not a claim that size wins on every matched panel.
+
+## Model-facing interpretation
+
+The models below look most like subgroup-scaffold preservers: they keep keys, row counts, and output structure more reliably on the size side than on the internal side.
+
+- `TVAE`: support advantage `0.035`, key advantage `0.049`. Reading: size-led scaffold preserver.
+- `TabPFGen`: support advantage `0.034`, key advantage `0.030`. Reading: size-led scaffold preserver.
+- `TabDDPM`: support advantage `0.027`, key advantage `0.044`. Reading: moderate size-led scaffold preserver.
+- `CTGAN`: support advantage `0.027`, key advantage `0.031`. Reading: moderate size-led scaffold preserver.
+
+These models look less like clean size-led preservers:
+
+- `ForestDiffusion`: support advantage `-0.011`, key advantage `-0.002`. Reading: no clear scaffold edge.
+- `ARF`: support advantage `-0.012`, key advantage `-0.001`. Reading: no clear scaffold edge.
+- `TabSyn`: support advantage `-0.020`, key advantage `0.032`. Reading: key retention without stable size scaffold.
+
+A particularly informative mixed case:
+
+- `TabSyn`: support advantage `-0.020` but key advantage `0.032`. Reading: subgroup identities are often retained, but the row-count / table-scaffold side is not stable enough to turn that into a size-led advantage.
+
+## What ability is this really measuring?
+
+- `subgroup_size_stability` on the shared basis is mostly a **subgroup scaffold preservation** ability: whether the same groups appear, whether rough prevalence is retained, and whether result-table structure survives.
+- `internal_profile_stability` asks for a harder **within-group analytical structure** ability: preserving ranking, filtered subgroup summaries, local two-dimensional geometry, and group-specific contrast after the subgroup has already been identified.
+
+## Prefix pattern
+
+- The clearest size-led regime is `m`, where the shared-basis support advantage is largest.
+- `c` is the main caution regime: on the shared support basis it is roughly balanced or slightly internal-favored, so the size story should not be overstated as universal.
+- `n` still shows a positive support-side edge, but much smaller than `m`.
+
+- `c`: support delta `-0.015`, key delta `-0.010`, `size > internal` share `0.322`.
+- `m`: support delta `0.032`, key delta `0.056`, `size > internal` share `0.626`.
+- `n`: support delta `0.024`, key delta `0.041`, `size > internal` share `0.417`.

evaluation/query_family/subgroup/shared_basis_stability_analysis/README.md

@@ -0,0 +1,22 @@
+# Shared-Basis Stability Analysis
+
+This folder builds figures for the claim that subgroup size is more stable
+than internal profile only after both branches are projected onto the same
+support-oriented basis.
+
+The shared metric used here is:
+
+- `support3_score = mean(key_set_score, row_count_score, column_score)`
+
+Outputs:
+
+- `support3_query_ecdf.png` / `.tex`
+- `support3_paired_delta_hist.png` / `.tex`
+- `support3_model_advantage_bars.png` / `.tex`
+- `support3_ecdf_curves.csv`
+- `support3_paired_delta_hist.csv`
+- `support3_model_advantage.csv`
+- `support3_prefix_advantage.csv`
+- `support3_paired_panel_table.csv`
+- `summary_metrics.json`
+- `INSIGHTS.md`

evaluation/query_family/subgroup/shared_basis_stability_analysis/build_shared_basis_stability_analysis.py

@@ -0,0 +1,636 @@
+from __future__ import annotations
+
+import json
+from pathlib import Path
+
+import matplotlib.pyplot as plt
+import numpy as np
+import pandas as pd
+
+
+RUN_TAG = "v2_keyset_cts_48_20260504_2350"
+INTERNAL_COLOR = "#AA3377"
+SIZE_COLOR = "#009988"
+MODEL_COLORS = {
+    "ARF": "#777777",
+    "BayesNet": "#CCBB44",
+    "CTGAN": "#EE6677",
+    "ForestDiffusion": "#228833",
+    "RealTabFormer": "#332288",
+    "TabbyFlow": "#882255",
+    "TabDDPM": "#EE7733",
+    "TabDiff": "#AA3377",
+    "TabPFGen": "#009988",
+    "TabSyn": "#66CCEE",
+    "TVAE": "#4477AA",
+}
+
+
+def load_query_rows(data_dir: Path) -> pd.DataFrame:
+    df = pd.read_csv(data_dir / "subgroup_query_rows.csv")
+    df = df[df["model_id"] != "real"].copy()
+    df["query_score"] = pd.to_numeric(df["query_score"], errors="coerce")
+    return df
+
+
+def load_query_details(run_dir: Path, sub_df: pd.DataFrame) -> pd.DataFrame:
+    needed = set(zip(sub_df["dataset_id"], sub_df["model_id"], sub_df["asset_key"], sub_df["query_id"]))
+    rows: list[dict[str, object]] = []
+
+    for dataset_id in sorted(sub_df["dataset_id"].dropna().unique()):
+        path = run_dir / "datasets" / str(dataset_id) / f"analysis_query_scores__{dataset_id}.jsonl"
+        with path.open("r", encoding="utf-8") as handle:
+            for line in handle:
+                obj = json.loads(line)
+                key = (obj["dataset_id"], obj["model_id"], obj["asset_key"], obj["query_id"])
+                if key not in needed:
+                    continue
+                details = obj.get("details") or {}
+                rows.append(
+                    {
+                        "dataset_id": obj["dataset_id"],
+                        "model_id": obj["model_id"],
+                        "asset_key": obj["asset_key"],
+                        "query_id": obj["query_id"],
+                        "strict_set_score": details.get("strict_set_score", details.get("set_score")),
+                        "key_set_score": details.get("key_set_score"),
+                        "profile_score": details.get("profile_score"),
+                        "row_count_score": details.get("row_count_score"),
+                        "column_score": details.get("column_score"),
+                    }
+                )
+
+    return pd.DataFrame(rows)
+
+
+def add_scores(df: pd.DataFrame) -> pd.DataFrame:
+    out = df.copy()
+    for col in [
+        "strict_set_score",
+        "key_set_score",
+        "profile_score",
+        "row_count_score",
+        "column_score",
+        "query_score",
+    ]:
+        out[col] = pd.to_numeric(out[col], errors="coerce")
+    out["support3_score"] = out[["key_set_score", "row_count_score", "column_score"]].mean(axis=1)
+    out["equal5_score"] = out[
+        ["strict_set_score", "key_set_score", "profile_score", "row_count_score", "column_score"]
+    ].mean(axis=1)
+    return out
+
+
+def ecdf(values: pd.Series) -> tuple[np.ndarray, np.ndarray]:
+    arr = np.sort(values.dropna().to_numpy(dtype=float))
+    if arr.size == 0:
+        return np.array([]), np.array([])
+    y = np.arange(1, arr.size + 1, dtype=float) / arr.size
+    return arr, y
+
+
+def save_ecdf_csv(internal: pd.Series, size: pd.Series, out_path: Path) -> pd.DataFrame:
+    grid = np.linspace(0.0, 1.0, 501)
+    i_vals = internal.dropna().to_numpy(dtype=float)
+    s_vals = size.dropna().to_numpy(dtype=float)
+    internal_ecdf = np.searchsorted(np.sort(i_vals), grid, side="right") / max(len(i_vals), 1)
+    size_ecdf = np.searchsorted(np.sort(s_vals), grid, side="right") / max(len(s_vals), 1)
+    df = pd.DataFrame(
+        {
+            "score": grid,
+            "internal_ecdf": internal_ecdf,
+            "size_ecdf": size_ecdf,
+            "internal_survival": 1.0 - internal_ecdf,
+            "size_survival": 1.0 - size_ecdf,
+        }
+    )
+    df.to_csv(out_path, index=False)
+    return df
+
+
+def save_delta_hist_csv(delta: pd.Series, out_path: Path) -> pd.DataFrame:
+    bins = np.linspace(-0.40, 0.40, 33)
+    hist, edges = np.histogram(delta.dropna().to_numpy(dtype=float), bins=bins, density=True)
+    centers = 0.5 * (edges[:-1] + edges[1:])
+    df = pd.DataFrame(
+        {
+            "bin_left": edges[:-1],
+            "bin_right": edges[1:],
+            "bin_center": centers,
+            "density": hist,
+        }
+    )
+    df.to_csv(out_path, index=False)
+    return df
+
+
+def build_panel_table(merged: pd.DataFrame) -> pd.DataFrame:
+    panel = (
+        merged.groupby(["dataset_prefix", "dataset_id", "model_id", "model_label", "subitem_id"], as_index=False)[
+            ["support3_score", "key_set_score", "query_score"]
+        ]
+        .mean()
+    )
+    wide = panel.pivot(
+        index=["dataset_prefix", "dataset_id", "model_id", "model_label"],
+        columns="subitem_id",
+        values=["support3_score", "key_set_score", "query_score"],
+    )
+    wide.columns = ["__".join(col).strip() for col in wide.columns.values]
+    wide = wide.reset_index()
+    wide = wide.dropna(
+        subset=[
+            "support3_score__internal_profile_stability",
+            "support3_score__subgroup_size_stability",
+        ]
+    ).copy()
+    wide["support3_delta_size_minus_internal"] = (
+        wide["support3_score__subgroup_size_stability"]
+        - wide["support3_score__internal_profile_stability"]
+    )
+    wide["key_delta_size_minus_internal"] = (
+        wide["key_set_score__subgroup_size_stability"]
+        - wide["key_set_score__internal_profile_stability"]
+    )
+    wide["query_delta_size_minus_internal"] = (
+        wide["query_score__subgroup_size_stability"]
+        - wide["query_score__internal_profile_stability"]
+    )
+    return wide
+
+
+def classify_model_ability(row: pd.Series) -> str:
+    support3_diff = float(row["support3_diff"])
+    key_diff = float(row["key_diff"])
+    if support3_diff >= 0.03:
+        return "Size-led scaffold preserver"
+    if support3_diff >= 0.015 and key_diff >= 0.02:
+        return "Moderate size-led scaffold preserver"
+    if support3_diff >= 0.0 and key_diff >= 0.015:
+        return "Near-balanced with mild size edge"
+    if support3_diff < 0.0 and key_diff > 0.02:
+        return "Key retention without stable size scaffold"
+    if support3_diff < 0.0 and key_diff <= 0.0:
+        return "No clear scaffold edge"
+    return "Mixed or weak separation"
+
+
+def build_model_summary(panel_wide: pd.DataFrame, out_path: Path) -> pd.DataFrame:
+    model = (
+        panel_wide.groupby(["model_id", "model_label"], as_index=False)
+        .agg(
+            paired_panel_count=("dataset_id", "size"),
+            support3_diff=("support3_delta_size_minus_internal", "mean"),
+            key_diff=("key_delta_size_minus_internal", "mean"),
+            query_diff=("query_delta_size_minus_internal", "mean"),
+            size_better_support3_share=("support3_delta_size_minus_internal", lambda s: (s > 0).mean()),
+        )
+        .sort_values("support3_diff", ascending=False)
+    )
+    model["ability_reading"] = model.apply(classify_model_ability, axis=1)
+    model.to_csv(out_path, index=False)
+    return model
+
+
+def build_prefix_summary(panel_wide: pd.DataFrame, out_path: Path) -> pd.DataFrame:
+    prefix = (
+        panel_wide.groupby("dataset_prefix", as_index=False)
+        .agg(
+            paired_panel_count=("dataset_id", "size"),
+            support3_diff=("support3_delta_size_minus_internal", "mean"),
+            key_diff=("key_delta_size_minus_internal", "mean"),
+            size_better_support3_share=("support3_delta_size_minus_internal", lambda s: (s > 0).mean()),
+        )
+        .sort_values("dataset_prefix")
+    )
+    prefix.to_csv(out_path, index=False)
+    return prefix
+
+
+def write_ecdf_png(ecdf_df: pd.DataFrame, internal: pd.Series, size: pd.Series, out_path: Path) -> None:
+    plt.style.use("seaborn-v0_8-whitegrid")
+    fig, ax = plt.subplots(figsize=(8.8, 5.4))
+    ax.plot(
+        ecdf_df["score"],
+        ecdf_df["internal_survival"],
+        color=INTERNAL_COLOR,
+        linewidth=2.4,
+        label="Internal profile stability",
+    )
+    ax.plot(
+        ecdf_df["score"],
+        ecdf_df["size_survival"],
+        color=SIZE_COLOR,
+        linewidth=2.4,
+        label="Subgroup size stability",
+    )
+    ax.axvline(float(internal.mean()), color=INTERNAL_COLOR, linestyle="--", linewidth=1.1, alpha=0.8)
+    ax.axvline(float(size.mean()), color=SIZE_COLOR, linestyle="--", linewidth=1.1, alpha=0.8)
+    ax.set_xlim(0.0, 1.0)
+    ax.set_ylim(0.0, 1.02)
+    ax.set_xlabel("Support3 score")
+    ax.set_ylabel("Share of queries with score >= x")
+    ax.set_title("Support-side stability on a shared subgroup basis")
+    ax.text(
+        0.015,
+        0.97,
+        (
+            f"Means: internal = {internal.mean():.3f}, size = {size.mean():.3f}\n"
+            f"Medians: internal = {internal.median():.3f}, size = {size.median():.3f}\n"
+            f"Low-tail share (<= 0.4): internal = {(internal <= 0.4).mean():.3f}, "
+            f"size = {(size <= 0.4).mean():.3f}"
+        ),
+        transform=ax.transAxes,
+        va="top",
+        ha="left",
+        fontsize=9.0,
+        bbox={"boxstyle": "round,pad=0.25", "facecolor": "white", "edgecolor": "#CCCCCC", "alpha": 0.96},
+    )
+    ax.legend(frameon=False, loc="upper right")
+    ax.grid(True, color="#D8DDE3", linewidth=0.8, alpha=0.8)
+    ax.spines["top"].set_visible(False)
+    ax.spines["right"].set_visible(False)
+    fig.tight_layout()
+    fig.savefig(out_path, dpi=240)
+    plt.close(fig)
+
+
+def write_delta_hist_png(panel_wide: pd.DataFrame, out_path: Path) -> None:
+    plt.style.use("seaborn-v0_8-whitegrid")
+    fig, ax = plt.subplots(figsize=(8.8, 5.4))
+    delta = panel_wide["support3_delta_size_minus_internal"].dropna()
+    ax.hist(delta, bins=np.linspace(-0.40, 0.40, 33), density=True, color=SIZE_COLOR, alpha=0.30, edgecolor="white")
+    ax.axvline(0.0, color="black", linestyle="--", linewidth=1.0, alpha=0.8)
+    ax.axvline(float(delta.mean()), color=SIZE_COLOR, linestyle="-", linewidth=2.0, alpha=0.9)
+    ax.set_xlim(-0.40, 0.40)
+    ax.set_xlabel("Panel-level support3(size) - support3(internal)")
+    ax.set_ylabel("Relative density")
+    ax.set_title("Matched-panel subgroup support advantage")
+    ax.text(
+        0.015,
+        0.97,
+        (
+            f"Panels: {len(delta)}\n"
+            f"Mean delta: {delta.mean():.3f}\n"
+            f"Share > 0: {(delta > 0).mean():.3f}\n"
+            f"Share >= 0.10: {(delta >= 0.10).mean():.3f}"
+        ),
+        transform=ax.transAxes,
+        va="top",
+        ha="left",
+        fontsize=9.0,
+        bbox={"boxstyle": "round,pad=0.25", "facecolor": "white", "edgecolor": "#CCCCCC", "alpha": 0.96},
+    )
+    ax.grid(True, color="#D8DDE3", linewidth=0.8, alpha=0.8)
+    ax.spines["top"].set_visible(False)
+    ax.spines["right"].set_visible(False)
+    fig.tight_layout()
+    fig.savefig(out_path, dpi=240)
+    plt.close(fig)
+
+
+def write_model_bar_png(model_df: pd.DataFrame, out_path: Path) -> None:
+    ordered = model_df.sort_values("support3_diff", ascending=False).copy()
+    colors = [MODEL_COLORS.get(label, "#666666") for label in ordered["model_label"]]
+    plt.style.use("seaborn-v0_8-whitegrid")
+    fig, ax = plt.subplots(figsize=(10.8, 5.8))
+    x = np.arange(len(ordered))
+    ax.bar(x, ordered["support3_diff"], color=colors, alpha=0.88)
+    ax.axhline(0.0, color="black", linestyle="--", linewidth=1.0, alpha=0.8)
+    ax.set_xticks(x)
+    ax.set_xticklabels(ordered["model_label"], rotation=40, ha="right")
+    ax.set_ylabel("Mean panel support3(size) - support3(internal)")
+    ax.set_title("Model-level scaffold-preservation advantage")
+    ax.grid(True, axis="y", color="#D8DDE3", linewidth=0.8, alpha=0.8)
+    ax.spines["top"].set_visible(False)
+    ax.spines["right"].set_visible(False)
+    for xi, (_, row) in enumerate(ordered.iterrows()):
+        ax.text(
+            xi,
+            row["support3_diff"] + (0.002 if row["support3_diff"] >= 0 else -0.006),
+            f"{row['support3_diff']:.3f}",
+            ha="center",
+            va="bottom" if row["support3_diff"] >= 0 else "top",
+            fontsize=8,
+        )
+    fig.tight_layout()
+    fig.savefig(out_path, dpi=240)
+    plt.close(fig)
+
+
+def write_ecdf_tex(ecdf_df: pd.DataFrame, internal: pd.Series, size: pd.Series, out_dir: Path) -> None:
+    ymax = 1.02
+    tex = rf"""\documentclass[tikz,border=4pt]{{standalone}}
+\usepackage{{pgfplots}}
+\pgfplotsset{{compat=1.18}}
+\definecolor{{internalcolor}}{{HTML}}{{AA3377}}
+\definecolor{{sizecolor}}{{HTML}}{{009988}}
+\begin{{document}}
+\begin{{tikzpicture}}
+\begin{{axis}}[
+width=14.8cm,
+height=8.7cm,
+xmin=0.0, xmax=1.0,
+ymin=0.0, ymax={ymax},
+xlabel={{Support3 score}},
+ylabel={{Share of queries with score >= x}},
+title={{Support-side stability on a shared subgroup basis}},
+ymajorgrids,
+grid style={{draw=gray!20}},
+major grid style={{draw=gray!30}},
+axis line style={{draw=black!70}},
+tick style={{draw=black!70}},
+legend style={{draw=none, fill=none, font=\small, at={{(0.98,0.98)}}, anchor=north east}},
+]
+\addplot[line width=1.9pt, color=internalcolor]
+table[x=score, y=internal_survival, col sep=comma]{{support3_ecdf_curves.csv}};
+\addlegendentry{{Internal profile stability}}
+\addplot[line width=1.9pt, color=sizecolor]
+table[x=score, y=size_survival, col sep=comma]{{support3_ecdf_curves.csv}};
+\addlegendentry{{Subgroup size stability}}
+\addplot[dashed, line width=1.0pt, color=internalcolor, opacity=0.85]
+coordinates {{({float(internal.mean()):.6f},0) ({float(internal.mean()):.6f},{ymax})}};
+\addplot[dashed, line width=1.0pt, color=sizecolor, opacity=0.85]
+coordinates {{({float(size.mean()):.6f},0) ({float(size.mean()):.6f},{ymax})}};
+\node[
+anchor=north west,
+draw=gray!55,
+fill=white,
+rounded corners=2pt,
+fill opacity=0.96,
+text opacity=1,
+align=left,
+font=\small
+] at (axis description cs:0.015,0.97) {{
+Means: internal = {float(internal.mean()):.3f}, size = {float(size.mean()):.3f}\\
+Medians: internal = {float(internal.median()):.3f}, size = {float(size.median()):.3f}\\
+Low-tail share (<= 0.4): internal = {float((internal <= 0.4).mean()):.3f}, size = {float((size <= 0.4).mean()):.3f}
+}};
+\end{{axis}}
+\end{{tikzpicture}}
+\end{{document}}
+"""
+    (out_dir / "support3_query_ecdf.tex").write_text(tex, encoding="utf-8")
+
+
+def write_delta_tex(hist_df: pd.DataFrame, panel_wide: pd.DataFrame, out_dir: Path) -> None:
+    ymax = float(hist_df["density"].max() * 1.10)
+    delta = panel_wide["support3_delta_size_minus_internal"].dropna()
+    tex = rf"""\documentclass[tikz,border=4pt]{{standalone}}
+\usepackage{{pgfplots}}
+\pgfplotsset{{compat=1.18}}
+\definecolor{{sizecolor}}{{HTML}}{{009988}}
+\begin{{document}}
+\begin{{tikzpicture}}
+\begin{{axis}}[
+width=14.8cm,
+height=8.7cm,
+xmin=-0.40, xmax=0.40,
+ymin=0.0, ymax={ymax:.4f},
+xlabel={{Panel-level support3(size) - support3(internal)}},
+ylabel={{Relative density}},
+title={{Matched-panel subgroup support advantage}},
+ymajorgrids,
+grid style={{draw=gray!20}},
+major grid style={{draw=gray!30}},
+axis line style={{draw=black!70}},
+tick style={{draw=black!70}},
+]
+\addplot[ybar interval, mark=none, draw=white, fill=sizecolor, fill opacity=0.30]
+table[x=bin_left, y=density, col sep=comma]{{support3_paired_delta_hist.csv}};
+\addplot[dashed, line width=1.0pt, color=black, opacity=0.85]
+coordinates {{(0,0) (0,{ymax:.4f})}};
+\addplot[line width=1.6pt, color=sizecolor]
+coordinates {{({float(delta.mean()):.6f},0) ({float(delta.mean()):.6f},{ymax:.4f})}};
+\node[
+anchor=north west,
+draw=gray!55,
+fill=white,
+rounded corners=2pt,
+fill opacity=0.96,
+text opacity=1,
+align=left,
+font=\small
+] at (axis description cs:0.015,0.97) {{
+Panels: {len(delta)}\\
+Mean delta: {float(delta.mean()):.3f}\\
+Share > 0: {float((delta > 0).mean()):.3f}\\
+Share >= 0.10: {float((delta >= 0.10).mean()):.3f}
+}};
+\end{{axis}}
+\end{{tikzpicture}}
+\end{{document}}
+"""
+    (out_dir / "support3_paired_delta_hist.tex").write_text(tex, encoding="utf-8")
+
+
+def write_model_bar_tex(model_df: pd.DataFrame, out_dir: Path) -> None:
+    ordered = model_df.sort_values("support3_diff", ascending=False).copy()
+    coords = " ".join(f"({row.model_label},{row.support3_diff:.6f})" for row in ordered.itertuples())
+    color_defs = "\n".join(
+        f"\\definecolor{{model{row.model_label.lower()}}}{{HTML}}{{{MODEL_COLORS.get(row.model_label, '#666666').lstrip('#')}}}"
+        for row in ordered.itertuples()
+    )
+    addplots = []
+    for row in ordered.itertuples():
+        addplots.append(
+            rf"\addplot+[ybar, bar width=7.5pt, draw=model{row.model_label.lower()}, fill=model{row.model_label.lower()}] coordinates {{ ({row.model_label},{row.support3_diff:.6f}) }};"
+        )
+    symbolic = ",".join(ordered["model_label"].tolist())
+    tex = rf"""\documentclass[tikz,border=4pt]{{standalone}}
+\usepackage{{pgfplots}}
+\pgfplotsset{{compat=1.18}}
+{color_defs}
+\begin{{document}}
+\begin{{tikzpicture}}
+\begin{{axis}}[
+width=17.0cm,
+height=8.7cm,
+ymin=-0.02, ymax={max(0.09, float(ordered['support3_diff'].max()) + 0.02):.4f},
+xlabel={{Model}},
+ylabel={{Mean panel support3(size) - support3(internal)}},
+title={{Model-level scaffold-preservation advantage}},
+ymajorgrids,
+grid style={{draw=gray!20}},
+major grid style={{draw=gray!30}},
+axis line style={{draw=black!70}},
+tick style={{draw=black!70}},
+symbolic x coords={{{symbolic}}},
+xtick=data,
+xticklabel style={{rotate=40, anchor=east, font=\scriptsize}},
+]
+\addplot[dashed, line width=1.0pt, color=black, opacity=0.85] coordinates {{({ordered.iloc[0]['model_label']},0) ({ordered.iloc[-1]['model_label']},0)}};
+{"".join(addplots)}
+\end{{axis}}
+\end{{tikzpicture}}
+\end{{document}}
+"""
+    (out_dir / "support3_model_advantage_bars.tex").write_text(tex, encoding="utf-8")
+
+
+def write_insights_md(
+    internal: pd.Series,
+    size: pd.Series,
+    panel_wide: pd.DataFrame,
+    model_df: pd.DataFrame,
+    prefix_df: pd.DataFrame,
+    out_path: Path,
+) -> None:
+    top_models = model_df.head(4)
+    bottom_models = model_df.tail(3)
+    lines = [
+        "# Shared-Basis Stability Insights",
+        "",
+        "## Main reading",
+        "",
+        (
+            "On the shared support-oriented basis `support3_score = mean(key_set_score, row_count_score, column_score)`, "
+            "subgroup-size queries are more robust than internal-profile queries."
+        ),
+        "",
+        f"- Query-level means: `internal = {float(internal.mean()):.3f}`, `size = {float(size.mean()):.3f}`",
+        f"- Query-level medians: `internal = {float(internal.median()):.3f}`, `size = {float(size.median()):.3f}`",
+        f"- Low-tail share `<= 0.4`: `internal = {float((internal <= 0.4).mean()):.3f}`, `size = {float((size <= 0.4).mean()):.3f}`",
+        f"- High-score share `>= 0.8`: `internal = {float((internal >= 0.8).mean()):.3f}`, `size = {float((size >= 0.8).mean()):.3f}`",
+        "",
+        "## Matched-panel interpretation",
+        "",
+        (
+            "Comparing the two subgroup branches inside the same dataset-model panel avoids raw-score incomparability. "
+            "Under this matched view, `support3(size) - support3(internal)` is positive on average."
+        ),
+        "",
+        f"- Mean matched-panel delta: `{float(panel_wide['support3_delta_size_minus_internal'].mean()):.3f}`",
+        f"- Share of panels with `size > internal`: `{float((panel_wide['support3_delta_size_minus_internal'] > 0).mean()):.3f}`",
+        f"- Share of panels with `size - internal >= 0.10`: `{float((panel_wide['support3_delta_size_minus_internal'] >= 0.10).mean()):.3f}`",
+        "- This is therefore a distribution-level stability story, not a claim that size wins on every matched panel.",
+        "",
+        "## Model-facing interpretation",
+        "",
+        "The models below look most like subgroup-scaffold preservers: they keep keys, row counts, and output structure more reliably on the size side than on the internal side.",
+        "",
+    ]
+    for row in top_models.itertuples():
+        lines.append(
+            f"- `{row.model_label}`: support advantage `{row.support3_diff:.3f}`, key advantage `{row.key_diff:.3f}`. "
+            f"Reading: {row.ability_reading.lower()}."
+        )
+    lines.extend(
+        [
+            "",
+            "These models look less like clean size-led preservers:",
+            "",
+        ]
+    )
+    for row in bottom_models.itertuples():
+        lines.append(
+            f"- `{row.model_label}`: support advantage `{row.support3_diff:.3f}`, key advantage `{row.key_diff:.3f}`. "
+            f"Reading: {row.ability_reading.lower()}."
+        )
+    tabsyn = model_df.loc[model_df["model_label"] == "TabSyn"]
+    if not tabsyn.empty:
+        row = tabsyn.iloc[0]
+        lines.extend(
+            [
+                "",
+                "A particularly informative mixed case:",
+                "",
+                (
+                    f"- `TabSyn`: support advantage `{row['support3_diff']:.3f}` but key advantage `{row['key_diff']:.3f}`. "
+                    "Reading: subgroup identities are often retained, but the row-count / table-scaffold side is not stable enough "
+                    "to turn that into a size-led advantage."
+                ),
+            ]
+        )
+    lines.extend(
+        [
+            "",
+            "## What ability is this really measuring?",
+            "",
+            "- `subgroup_size_stability` on the shared basis is mostly a **subgroup scaffold preservation** ability: whether the same groups appear, whether rough prevalence is retained, and whether result-table structure survives.",
+            "- `internal_profile_stability` asks for a harder **within-group analytical structure** ability: preserving ranking, filtered subgroup summaries, local two-dimensional geometry, and group-specific contrast after the subgroup has already been identified.",
+            "",
+            "## Prefix pattern",
+            "",
+            "- The clearest size-led regime is `m`, where the shared-basis support advantage is largest.",
+            "- `c` is the main caution regime: on the shared support basis it is roughly balanced or slightly internal-favored, so the size story should not be overstated as universal.",
+            "- `n` still shows a positive support-side edge, but much smaller than `m`.",
+            "",
+        ]
+    )
+    for row in prefix_df.itertuples():
+        lines.append(
+            f"- `{row.dataset_prefix}`: support delta `{row.support3_diff:.3f}`, key delta `{row.key_diff:.3f}`, "
+            f"`size > internal` share `{row.size_better_support3_share:.3f}`."
+        )
+    out_path.write_text("\n".join(lines) + "\n", encoding="utf-8")
+
+
+def main() -> None:
+    out_dir = Path(__file__).resolve().parent
+    project_root = Path(__file__).resolve().parents[4]
+    subgroup_root = project_root / "Evaluation" / "query_fivepart_breakdown" / "subgroup_breakdown"
+    run_dir = project_root / "Evaluation" / "analysis" / "runs" / RUN_TAG
+
+    query_rows = load_query_rows(subgroup_root / "data")
+    detail_rows = load_query_details(run_dir, query_rows)
+    merged = query_rows.merge(
+        detail_rows,
+        on=["dataset_id", "model_id", "asset_key", "query_id"],
+        how="left",
+    )
+    merged = add_scores(merged)
+
+    internal = merged.loc[
+        merged["subitem_id"] == "internal_profile_stability", "support3_score"
+    ].astype(float)
+    size = merged.loc[
+        merged["subitem_id"] == "subgroup_size_stability", "support3_score"
+    ].astype(float)
+
+    ecdf_df = save_ecdf_csv(internal, size, out_dir / "support3_ecdf_curves.csv")
+    panel_wide = build_panel_table(merged)
+    hist_df = save_delta_hist_csv(
+        panel_wide["support3_delta_size_minus_internal"],
+        out_dir / "support3_paired_delta_hist.csv",
+    )
+    model_df = build_model_summary(panel_wide, out_dir / "support3_model_advantage.csv")
+    prefix_df = build_prefix_summary(panel_wide, out_dir / "support3_prefix_advantage.csv")
+    panel_wide.to_csv(out_dir / "support3_paired_panel_table.csv", index=False)
+
+    summary = {
+        "run_tag": RUN_TAG,
+        "metric": "support3_score",
+        "definition": "mean(key_set_score, row_count_score, column_score)",
+        "query_level": {
+            "internal_mean": float(internal.mean()),
+            "size_mean": float(size.mean()),
+            "internal_median": float(internal.median()),
+            "size_median": float(size.median()),
+            "internal_share_le_0_4": float((internal <= 0.4).mean()),
+            "size_share_le_0_4": float((size <= 0.4).mean()),
+            "internal_share_ge_0_8": float((internal >= 0.8).mean()),
+            "size_share_ge_0_8": float((size >= 0.8).mean()),
+        },
+        "panel_level": {
+            "paired_panel_count": int(len(panel_wide)),
+            "mean_delta": float(panel_wide["support3_delta_size_minus_internal"].mean()),
+            "size_better_share": float((panel_wide["support3_delta_size_minus_internal"] > 0).mean()),
+            "size_advantage_ge_0_10_share": float((panel_wide["support3_delta_size_minus_internal"] >= 0.10).mean()),
+        },
+    }
+    (out_dir / "summary_metrics.json").write_text(json.dumps(summary, indent=2, ensure_ascii=False), encoding="utf-8")
+
+    write_ecdf_png(ecdf_df, internal, size, out_dir / "support3_query_ecdf.png")
+    write_delta_hist_png(panel_wide, out_dir / "support3_paired_delta_hist.png")
+    write_model_bar_png(model_df, out_dir / "support3_model_advantage_bars.png")
+
+    write_ecdf_tex(ecdf_df, internal, size, out_dir)
+    write_delta_tex(hist_df, panel_wide, out_dir)
+    write_model_bar_tex(model_df, out_dir)
+
+    write_insights_md(internal, size, panel_wide, model_df, prefix_df, out_dir / "INSIGHTS.md")
+
+
+if __name__ == "__main__":
+    main()

evaluation/query_family/subgroup/shared_basis_stability_analysis/summary_metrics.json

@@ -0,0 +1,21 @@
+{
+  "run_tag": "v2_keyset_cts_48_20260504_2350",
+  "metric": "support3_score",
+  "definition": "mean(key_set_score, row_count_score, column_score)",
+  "query_level": {
+    "internal_mean": 0.7770301021787321,
+    "size_mean": 0.8184448182874395,
+    "internal_median": 0.8888888888888888,
+    "size_median": 1.0,
+    "internal_share_le_0_4": 0.14303930393039305,
+    "size_share_le_0_4": 0.1292771599657827,
+    "internal_share_ge_0_8": 0.5377287728772877,
+    "size_share_ge_0_8": 0.6319503849443969
+  },
+  "panel_level": {
+    "paired_panel_count": 398,
+    "mean_delta": 0.011131230358720141,
+    "size_better_share": 0.4371859296482412,
+    "size_advantage_ge_0_10_share": 0.12562814070351758
+  }
+}

evaluation/query_family/subgroup/shared_basis_stability_analysis/support3_ecdf_curves.csv

@@ -0,0 +1,502 @@
+score,internal_ecdf,size_ecdf,internal_survival,size_survival
+0.0,0.0,0.0,1.0,1.0
+0.002,0.0,0.0,1.0,1.0
+0.004,0.0,0.0,1.0,1.0
+0.006,0.0,0.0,1.0,1.0
+0.008,0.0,0.0,1.0,1.0
+0.01,0.0,0.0,1.0,1.0
+0.012,0.0,0.0,1.0,1.0
+0.014,0.0,0.0,1.0,1.0
+0.016,0.0,0.0,1.0,1.0
+0.018000000000000002,0.0,0.0,1.0,1.0
+0.02,0.0,0.0,1.0,1.0
+0.022,0.0,0.0,1.0,1.0
+0.024,0.0,0.0,1.0,1.0
+0.026000000000000002,0.0,0.0,1.0,1.0
+0.028,0.0,0.0,1.0,1.0
+0.03,0.0,0.0,1.0,1.0
+0.032,0.0,0.0,1.0,1.0
+0.034,0.0,0.0,1.0,1.0
+0.036000000000000004,0.0,0.0,1.0,1.0
+0.038,0.0,0.0,1.0,1.0
+0.04,0.0,0.0,1.0,1.0
+0.042,0.0,0.0,1.0,1.0
+0.044,0.0,0.0,1.0,1.0
+0.046,0.0,0.0,1.0,1.0
+0.048,0.0,0.0,1.0,1.0
+0.05,0.0,0.0,1.0,1.0
+0.052000000000000005,0.0,0.0,1.0,1.0
+0.054,0.0,0.0,1.0,1.0
+0.056,0.0,0.0,1.0,1.0
+0.058,0.0,0.0,1.0,1.0
+0.06,0.0,0.0,1.0,1.0
+0.062,0.0,0.0,1.0,1.0
+0.064,0.0,0.0,1.0,1.0
+0.066,0.0,0.0,1.0,1.0
+0.068,0.0,0.0,1.0,1.0
+0.07,0.0,0.0,1.0,1.0
+0.07200000000000001,0.0,0.0,1.0,1.0
+0.074,0.0,0.0,1.0,1.0
+0.076,0.0,0.0,1.0,1.0
+0.078,0.0,0.0,1.0,1.0
+0.08,0.0,0.0,1.0,1.0
+0.082,0.0,0.0,1.0,1.0
+0.084,0.0,0.0,1.0,1.0
+0.08600000000000001,0.0,0.0,1.0,1.0
+0.088,0.0,0.0,1.0,1.0
+0.09,0.0,0.0,1.0,1.0
+0.092,0.0,0.0,1.0,1.0
+0.094,0.0,0.0,1.0,1.0
+0.096,0.0,0.0,1.0,1.0
+0.098,0.0,0.0,1.0,1.0
+0.1,0.0,0.0,1.0,1.0
+0.10200000000000001,0.0,0.0,1.0,1.0
+0.10400000000000001,0.0,0.0,1.0,1.0
+0.106,0.0,0.0,1.0,1.0
+0.108,0.0,0.0,1.0,1.0
+0.11,0.0,0.0,1.0,1.0
+0.112,0.0,0.0,1.0,1.0
+0.114,0.0,0.0,1.0,1.0
+0.116,0.0,0.0,1.0,1.0
+0.11800000000000001,0.0,0.0,1.0,1.0
+0.12,0.0,0.0,1.0,1.0
+0.122,0.0,0.0,1.0,1.0
+0.124,0.0,0.0,1.0,1.0
+0.126,0.0,0.0,1.0,1.0
+0.128,0.0,0.0,1.0,1.0
+0.13,0.0,0.0,1.0,1.0
+0.132,0.0,0.0,1.0,1.0
+0.134,0.0,0.0,1.0,1.0
+0.136,0.0,0.0,1.0,1.0
+0.138,0.0,0.0,1.0,1.0
+0.14,0.0,0.0,1.0,1.0
+0.14200000000000002,0.0,0.0,1.0,1.0
+0.14400000000000002,0.0,0.0,1.0,1.0
+0.146,0.0,0.0,1.0,1.0
+0.148,0.0,0.0,1.0,1.0
+0.15,0.0,0.0,1.0,1.0
+0.152,0.0,0.0,1.0,1.0
+0.154,0.0,0.0,1.0,1.0
+0.156,0.0,0.0,1.0,1.0
+0.158,0.0,0.0,1.0,1.0
+0.16,0.0,0.0,1.0,1.0
+0.162,0.0,0.0,1.0,1.0
+0.164,0.0,0.0,1.0,1.0
+0.166,0.0,0.0,1.0,1.0
+0.168,0.0,0.0,1.0,1.0
+0.17,0.0,0.0,1.0,1.0
+0.17200000000000001,0.0,0.0,1.0,1.0
+0.17400000000000002,0.0,0.0,1.0,1.0
+0.176,0.0,0.0,1.0,1.0
+0.178,0.0,0.0,1.0,1.0
+0.18,0.0,0.0,1.0,1.0
+0.182,0.0,0.0,1.0,1.0
+0.184,0.0,0.0,1.0,1.0
+0.186,0.0,0.0,1.0,1.0
+0.188,0.0,0.0,1.0,1.0
+0.19,0.0,0.0,1.0,1.0
+0.192,0.0,0.0,1.0,1.0
+0.194,0.0,0.0,1.0,1.0
+0.196,0.0,0.0,1.0,1.0
+0.198,0.0,0.0,1.0,1.0
+0.2,0.0,0.0,1.0,1.0
+0.202,0.0,0.0,1.0,1.0
+0.20400000000000001,0.0,0.0,1.0,1.0
+0.20600000000000002,0.0,0.0,1.0,1.0
+0.20800000000000002,0.0,0.0,1.0,1.0
+0.21,0.0,0.0,1.0,1.0
+0.212,0.0,0.0,1.0,1.0
+0.214,0.0,0.0,1.0,1.0
+0.216,0.0,0.0,1.0,1.0
+0.218,0.0,0.0,1.0,1.0
+0.22,0.0,0.0,1.0,1.0
+0.222,0.0,0.0,1.0,1.0
+0.224,0.0,0.0,1.0,1.0
+0.226,0.0,0.0,1.0,1.0
+0.228,0.0,0.0,1.0,1.0
+0.23,0.0,0.0,1.0,1.0
+0.232,0.0,0.0,1.0,1.0
+0.234,0.0,0.0,1.0,1.0
+0.23600000000000002,0.0,0.0,1.0,1.0
+0.23800000000000002,0.0,0.0,1.0,1.0
+0.24,0.0,0.0,1.0,1.0
+0.242,0.0,0.0,1.0,1.0
+0.244,0.0,0.0,1.0,1.0
+0.246,0.0,0.0,1.0,1.0
+0.248,0.0,0.0,1.0,1.0
+0.25,0.0,0.0,1.0,1.0
+0.252,0.0,0.0,1.0,1.0
+0.254,0.0,0.0,1.0,1.0
+0.256,0.0,0.0,1.0,1.0
+0.258,0.0,0.0,1.0,1.0
+0.26,0.0,0.0,1.0,1.0
+0.262,0.0,0.0,1.0,1.0
+0.264,0.0,0.0,1.0,1.0
+0.266,0.0,0.0,1.0,1.0
+0.268,0.0,0.0,1.0,1.0
+0.27,0.0,0.0,1.0,1.0
+0.272,0.0,0.0,1.0,1.0
+0.274,0.0,0.0,1.0,1.0
+0.276,0.0,0.0,1.0,1.0
+0.278,0.0,0.0,1.0,1.0
+0.28,0.0,0.0,1.0,1.0
+0.28200000000000003,0.0,0.0,1.0,1.0
+0.28400000000000003,0.0,0.0,1.0,1.0
+0.28600000000000003,0.0,0.0,1.0,1.0
+0.28800000000000003,0.0,0.0,1.0,1.0
+0.29,0.0,0.0,1.0,1.0
+0.292,0.0,0.0,1.0,1.0
+0.294,0.0,0.0,1.0,1.0
+0.296,0.0,0.0,1.0,1.0
+0.298,0.0,0.0,1.0,1.0
+0.3,0.0,0.0,1.0,1.0
+0.302,0.0,0.0,1.0,1.0
+0.304,0.0,0.0,1.0,1.0
+0.306,0.0,0.0,1.0,1.0
+0.308,0.0,0.0,1.0,1.0
+0.31,0.0,0.0,1.0,1.0
+0.312,0.0,0.0,1.0,1.0
+0.314,0.0,0.0,1.0,1.0
+0.316,0.0,0.0,1.0,1.0
+0.318,0.0,0.0,1.0,1.0
+0.32,0.0,0.0,1.0,1.0
+0.322,0.0,0.0,1.0,1.0
+0.324,0.0,0.0,1.0,1.0
+0.326,0.0,0.0,1.0,1.0
+0.328,0.0,0.0,1.0,1.0
+0.33,0.0,0.0,1.0,1.0
+0.332,0.0,0.0,1.0,1.0
+0.334,0.08620162153893224,0.07494623655913979,0.9137983784610677,0.9250537634408602
+0.336,0.08987303044209882,0.08043010752688172,0.9101269695579012,0.9195698924731183
+0.338,0.09201468563561267,0.0867741935483871,0.9079853143643873,0.9132258064516129
+0.34,0.09943399112742848,0.09419354838709677,0.9005660088725715,0.9058064516129032
+0.342,0.10364081382897354,0.09602150537634409,0.8963591861710265,0.9039784946236559
+0.34400000000000003,0.10540003059507419,0.09709677419354838,0.8945999694049258,0.9029032258064517
+0.34600000000000003,0.1066238335627964,0.09870967741935484,0.8933761664372036,0.9012903225806451
+0.34800000000000003,0.10983631635306716,0.09956989247311827,0.8901636836469329,0.9004301075268817
+0.35000000000000003,0.11090714394982408,0.10021505376344086,0.889092856050176,0.8997849462365591
+0.352,0.11266636071592473,0.10096774193548387,0.8873336392840753,0.8990322580645161
+0.354,0.11358421294171639,0.10225806451612904,0.8864157870582836,0.8977419354838709
+0.356,0.1156493804497476,0.10387096774193548,0.8843506195502524,0.8961290322580645
+0.358,0.11748508490133089,0.10559139784946237,0.8825149150986691,0.8944086021505376
+0.36,0.11947376472387945,0.10763440860215054,0.8805262352761205,0.8923655913978494
+0.362,0.12123298148998012,0.109247311827957,0.8787670185100199,0.890752688172043
+0.364,0.12192137065932385,0.10956989247311828,0.8780786293406762,0.8904301075268817
+0.366,0.12375707511090714,0.11053763440860215,0.8762429248890928,0.8894623655913978
+0.368,0.1249043903931467,0.11182795698924732,0.8750956096068533,0.8881720430107527
+0.37,0.12704604558666055,0.11279569892473118,0.8729539544133395,0.8872043010752688
+0.372,0.12895823772372647,0.11365591397849463,0.8710417622762735,0.8863440860215054
+0.374,0.12918770078017439,0.11526881720430107,0.8708122992198256,0.884731182795699
+0.376,0.1302585283769313,0.11645161290322581,0.8697414716230687,0.8835483870967742
+0.378,0.13171179440110142,0.11741935483870967,0.8682882055988985,0.8825806451612903
+0.38,0.13217072051399725,0.11741935483870967,0.8678292794860027,0.8825806451612903
+0.382,0.13362398653816734,0.12,0.8663760134618327,0.88
+0.384,0.1341594003365458,0.12,0.8658405996634542,0.88
+0.386,0.13645403090102493,0.12129032258064516,0.8635459690989751,0.8787096774193548
+0.388,0.1366070062719902,0.12129032258064516,0.8633929937280098,0.8787096774193548
+0.39,0.1385956860945388,0.12473118279569892,0.8614043139054612,0.875268817204301
+0.392,0.1388251491509867,0.12473118279569892,0.8611748508490134,0.875268817204301
+0.394,0.1408903166590179,0.1267741935483871,0.8591096833409821,0.873225806451613
+0.396,0.14142573045739637,0.12698924731182795,0.8585742695426036,0.8730107526881721
+0.398,0.14188465657029217,0.12698924731182795,0.8581153434297079,0.8730107526881721
+0.4,0.14586201621538933,0.13,0.8541379837846107,0.87
+0.402,0.14632094232828516,0.13010752688172042,0.8536790576717148,0.8698924731182796
+0.404,0.14785069603793788,0.13075268817204302,0.8521493039620621,0.869247311827957
+0.406,0.14853908520728162,0.1313978494623656,0.8514609147927183,0.8686021505376345
+0.40800000000000003,0.15182805568303503,0.13279569892473117,0.848171944316965,0.8672043010752688
+0.41000000000000003,0.15251644485237878,0.13279569892473117,0.8474835551476212,0.8672043010752688
+0.41200000000000003,0.15465810004589262,0.13365591397849463,0.8453418999541074,0.8663440860215054
+0.41400000000000003,0.15496405078782316,0.13365591397849463,0.8450359492121768,0.8663440860215054
+0.41600000000000004,0.15580541532813216,0.13376344086021505,0.8441945846718678,0.8662365591397849
+0.418,0.15848248432002449,0.1350537634408602,0.8415175156799755,0.8649462365591398
+0.42,0.1587119473764724,0.13548387096774195,0.8412880526235276,0.864516129032258
+0.422,0.1591708734893682,0.13602150537634408,0.8408291265106318,0.863978494623656
+0.424,0.1605476518280557,0.13688172043010752,0.8394523481719443,0.8631182795698925
+0.426,0.1613890163683647,0.13709677419354838,0.8386109836316353,0.8629032258064516
+0.428,0.16154199173932995,0.13709677419354838,0.8384580082606701,0.8629032258064516
+0.43,0.1671255927795625,0.139247311827957,0.8328744072204375,0.860752688172043
+0.432,0.16727856815052777,0.1396774193548387,0.8327214318494722,0.8603225806451613
+0.434,0.1681964203763194,0.140752688172043,0.8318035796236806,0.859247311827957
+0.436,0.17072051399724644,0.14225806451612905,0.8292794860027536,0.8577419354838709
+0.438,0.17117944011014227,0.1424731182795699,0.8288205598898577,0.8575268817204301
+0.44,0.17194431696496865,0.14268817204301076,0.8280556830350314,0.8573118279569892
+0.442,0.17263270613431236,0.14290322580645162,0.8273672938656876,0.8570967741935483
+0.444,0.17263270613431236,0.14290322580645162,0.8273672938656876,0.8570967741935483
+0.446,0.1784457702309928,0.14655913978494622,0.8215542297690072,0.8534408602150538
+0.448,0.1784457702309928,0.14655913978494622,0.8215542297690072,0.8534408602150538
+0.45,0.17890469634388864,0.14677419354838708,0.8210953036561114,0.853225806451613
+0.452,0.17966957319871502,0.14720430107526883,0.820330426801285,0.8527956989247312
+0.454,0.18181122839222885,0.1475268817204301,0.8181887716077711,0.85247311827957
+0.456,0.18318800673091631,0.14849462365591398,0.8168119932690837,0.8515053763440861
+0.458,0.18357044515832951,0.14870967741935484,0.8164295548416705,0.8512903225806452
+0.46,0.18379990821477743,0.14881720430107526,0.8162000917852226,0.8511827956989247
+0.462,0.18464127275508643,0.14913978494623656,0.8153587272449135,0.8508602150537634
+0.464,0.18479424812605172,0.14956989247311828,0.8152057518739483,0.8504301075268818
+0.466,0.18487073581153435,0.14956989247311828,0.8151292641884657,0.8504301075268818
+0.468,0.18762429248890927,0.15161290322580645,0.8123757075110907,0.8483870967741935
+0.47000000000000003,0.18800673091632247,0.1517204301075269,0.8119932690836775,0.8482795698924731
+0.47200000000000003,0.1884656570292183,0.15204301075268817,0.8115343429707818,0.8479569892473118
+0.47400000000000003,0.18953648462597522,0.15311827956989246,0.8104635153740247,0.8468817204301076
+0.47600000000000003,0.1899189230533884,0.15311827956989246,0.8100810769466116,0.8468817204301076
+0.47800000000000004,0.19045433685176685,0.15344086021505376,0.8095456631482332,0.8465591397849462
+0.48,0.1905308245372495,0.15354838709677418,0.8094691754627505,0.8464516129032258
+0.482,0.19121921370659323,0.1543010752688172,0.8087807862934068,0.8456989247311828
+0.484,0.19121921370659323,0.1543010752688172,0.8087807862934068,0.8456989247311828
+0.486,0.19160165213400643,0.15451612903225806,0.8083983478659935,0.8454838709677419
+0.488,0.19206057824690226,0.1546236559139785,0.8079394217530977,0.8453763440860215
+0.49,0.19267247973076335,0.15559139784946235,0.8073275202692367,0.8444086021505377
+0.492,0.19328438121462443,0.15698924731182795,0.8067156187853756,0.843010752688172
+0.494,0.19374330732752026,0.1570967741935484,0.8062566926724797,0.8429032258064516
+0.496,0.19374330732752026,0.15731182795698925,0.8062566926724797,0.8426881720430107
+0.498,0.1938197950130029,0.15731182795698925,0.806180204986997,0.8426881720430107
+0.5,0.2093467951659783,0.1664516129032258,0.7906532048340217,0.8335483870967741
+0.502,0.20957625822242618,0.1664516129032258,0.7904237417775738,0.8335483870967741
+0.504,0.20988220896435675,0.16666666666666666,0.7901177910356433,0.8333333333333334
+0.506,0.21011167202080466,0.16666666666666666,0.7898883279791953,0.8333333333333334
+0.508,0.2104176227627352,0.16666666666666666,0.7895823772372648,0.8333333333333334
+0.51,0.21148845035949213,0.16752688172043012,0.7885115496405078,0.8324731182795699
+0.512,0.21248279027076641,0.16774193548387098,0.7875172097292336,0.832258064516129
+0.514,0.2128652286981796,0.16795698924731184,0.7871347713018204,0.8320430107526882
+0.516,0.21416551935138442,0.1682795698924731,0.7858344806486156,0.8317204301075269
+0.518,0.21470093314976288,0.16849462365591397,0.7852990668502371,0.8315053763440861
+0.52,0.21500688389169345,0.16860215053763442,0.7849931161083066,0.8313978494623656
+0.522,0.21523634694814134,0.1689247311827957,0.7847636530518587,0.8310752688172043
+0.524,0.21668961297231146,0.16989247311827957,0.7833103870276885,0.8301075268817204
+0.526,0.21668961297231146,0.16989247311827957,0.7833103870276885,0.8301075268817204
+0.528,0.216995563714242,0.17043010752688173,0.783004436285758,0.8295698924731183
+0.53,0.2171485390852073,0.17096774193548386,0.7828514609147927,0.8290322580645162
+0.532,0.21722502677068992,0.17096774193548386,0.7827749732293101,0.8290322580645162
+0.534,0.2179899036255163,0.17279569892473118,0.7820100963744837,0.8272043010752688
+0.536,0.2182193666819642,0.17301075268817204,0.7817806333180358,0.826989247311828
+0.538,0.21844882973841212,0.17311827956989248,0.7815511702615878,0.8268817204301075
+0.54,0.21898424353679058,0.17344086021505375,0.7810157564632094,0.8265591397849462
+0.542,0.21951965733516904,0.17397849462365592,0.780480342664831,0.8260215053763441
+0.544,0.21982560807709958,0.1743010752688172,0.7801743919229004,0.8256989247311828
+0.546,0.2202845341899954,0.1746236559139785,0.7797154658100046,0.8253763440860216
+0.548,0.2204375095609607,0.17483870967741935,0.7795624904390392,0.8251612903225807
+0.55,0.22120238641578704,0.1750537634408602,0.778797613584213,0.8249462365591398
+0.552,0.22143184947223496,0.1750537634408602,0.7785681505277651,0.8249462365591398
+0.554,0.22158482484320025,0.1750537634408602,0.7784151751567998,0.8249462365591398
+0.556,0.22877466727856816,0.17849462365591398,0.7712253327214318,0.821505376344086
+0.558,0.22885115496405078,0.17849462365591398,0.7711488450359492,0.821505376344086
+0.56,0.22931008107694661,0.17849462365591398,0.7706899189230534,0.821505376344086
+0.562,0.22976900718984244,0.179247311827957,0.7702309928101576,0.820752688172043
+0.5640000000000001,0.2299219825608077,0.17956989247311828,0.7700780174391924,0.8204301075268817
+0.5660000000000001,0.23022793330273825,0.17989247311827958,0.7697720666972617,0.8201075268817204
+0.5680000000000001,0.23053388404466882,0.1803225806451613,0.7694661159553312,0.8196774193548387
+0.5700000000000001,0.23076334710111673,0.18053763440860215,0.7692366528988832,0.8194623655913978
+0.5720000000000001,0.23145173627046045,0.18064516129032257,0.7685482637295395,0.8193548387096774
+0.5740000000000001,0.23168119932690837,0.18064516129032257,0.7683188006730917,0.8193548387096774
+0.5760000000000001,0.23259905155270003,0.18193548387096775,0.7674009484473,0.8180645161290323
+0.578,0.2332109530365611,0.18204301075268817,0.7667890469634389,0.8179569892473119
+0.58,0.2333639284075264,0.18204301075268817,0.7666360715924736,0.8179569892473119
+0.582,0.23374636683493957,0.1821505376344086,0.7662536331650605,0.8178494623655914
+0.584,0.23535260823007495,0.18301075268817205,0.764647391769925,0.816989247311828
+0.586,0.23573504665748815,0.18301075268817205,0.7642649533425119,0.816989247311828
+0.588,0.23611748508490132,0.18301075268817205,0.7638825149150987,0.816989247311828
+0.59,0.23665289888327978,0.18408602150537634,0.7633471011167202,0.8159139784946237
+0.592,0.23703533731069298,0.18408602150537634,0.762964662689307,0.8159139784946237
+0.594,0.23764723879455407,0.1846236559139785,0.7623527612054459,0.8153763440860216
+0.596,0.23802967722196727,0.18548387096774194,0.7619703227780328,0.814516129032258
+0.598,0.23833562796389782,0.18623655913978496,0.7616643720361022,0.813763440860215
+0.6,0.2421600122380297,0.1881720430107527,0.7578399877619704,0.8118279569892473
+0.602,0.2423894752944776,0.1881720430107527,0.7576105247055224,0.8118279569892473
+0.604,0.24246596297996023,0.18838709677419355,0.7575340370200397,0.8116129032258065
+0.606,0.24315435214930395,0.18913978494623657,0.7568456478506961,0.8108602150537634
+0.608,0.24330732752026923,0.19010752688172042,0.7566926724797307,0.8098924731182796
+0.61,0.24376625363316506,0.19010752688172042,0.7562337463668349,0.8098924731182796
+0.612,0.24583142114119627,0.19118279569892474,0.7541685788588037,0.8088172043010753
+0.614,0.24598439651216153,0.1913978494623656,0.7540156034878385,0.8086021505376344
+0.616,0.24613737188312682,0.19172043010752687,0.7538626281168732,0.8082795698924732
+0.618,0.24636683493957473,0.19182795698924732,0.7536331650604253,0.8081720430107526
+0.62,0.2483555147621233,0.19301075268817205,0.7516444852378767,0.806989247311828
+0.622,0.24912039161694968,0.19311827956989247,0.7508796083830503,0.8068817204301075
+0.624,0.24950283004436286,0.19440860215053762,0.7504971699556371,0.8055913978494624
+0.626,0.2506501453266024,0.19473118279569893,0.7493498546733977,0.8052688172043011
+0.628,0.25072663301208503,0.19494623655913978,0.749273366987915,0.8050537634408602
+0.63,0.25195043597980726,0.1953763440860215,0.7480495640201927,0.8046236559139786
+0.632,0.2521034113507725,0.19559139784946236,0.7478965886492275,0.8044086021505377
+0.634,0.25256233746366835,0.19591397849462366,0.7474376625363317,0.8040860215053763
+0.636,0.2528682882055989,0.19602150537634408,0.7471317117944011,0.8039784946236559
+0.638,0.2530212635765642,0.19602150537634408,0.7469787364234358,0.8039784946236559
+0.64,0.2553158941410433,0.19655913978494624,0.7446841058589567,0.8034408602150538
+0.642,0.2561572586813523,0.19666666666666666,0.7438427413186477,0.8033333333333333
+0.644,0.2564632094232829,0.19688172043010752,0.7435367905767172,0.8031182795698925
+0.646,0.25791647544745294,0.19763440860215054,0.742083524552547,0.8023655913978495
+0.648,0.25868135230227934,0.1978494623655914,0.7413186476977207,0.8021505376344086
+0.65,0.2598286675845189,0.19795698924731184,0.740171332415481,0.8020430107526881
+0.652,0.2600581306409668,0.19795698924731184,0.7399418693590332,0.8020430107526881
+0.654,0.26044056906838,0.19806451612903225,0.73955943093162,0.8019354838709678
+0.656,0.26135842129417164,0.19806451612903225,0.7386415787058284,0.8019354838709678
+0.658,0.2618938350925501,0.1981720430107527,0.7381061649074498,0.8018279569892472
+0.66,0.26219978583448067,0.19838709677419356,0.7378002141655193,0.8016129032258065
+0.662,0.26235276120544593,0.19870967741935483,0.7376472387945541,0.8012903225806451
+0.664,0.26242924889092856,0.1989247311827957,0.7375707511090714,0.8010752688172043
+0.666,0.26273519963285913,0.1989247311827957,0.7372648003671409,0.8010752688172043
+0.668,0.3937586048646168,0.3135483870967742,0.6062413951353832,0.6864516129032259
+0.67,0.3938350925500994,0.3139784946236559,0.6061649074499006,0.6860215053763441
+0.672,0.3942175309775126,0.3139784946236559,0.6057824690224873,0.6860215053763441
+0.674,0.3943705063484779,0.3141935483870968,0.6056294936515221,0.6858064516129032
+0.676,0.39475294477589107,0.31473118279569895,0.6052470552241089,0.6852688172043011
+0.678,0.3962062108000612,0.3164516129032258,0.6037937891999388,0.6835483870967742
+0.68,0.3982713783080924,0.3190322580645161,0.6017286216919076,0.6809677419354838
+0.682,0.39918923053388405,0.32,0.600810769466116,0.6799999999999999
+0.684,0.4005660088725715,0.3210752688172043,0.5994339911274285,0.6789247311827957
+0.686,0.40148386109836315,0.3227956989247312,0.5985161389016369,0.6772043010752689
+0.6880000000000001,0.40247820100963744,0.3235483870967742,0.5975217989903625,0.6764516129032259
+0.6900000000000001,0.4033960532354291,0.32365591397849464,0.596603946764571,0.6763440860215053
+0.6920000000000001,0.40354902860639436,0.32408602150537635,0.5964509713936057,0.6759139784946236
+0.6940000000000001,0.4040079547192902,0.3243010752688172,0.5959920452807098,0.6756989247311829
+0.6960000000000001,0.4046198562031513,0.32505376344086023,0.5953801437968487,0.6749462365591398
+0.6980000000000001,0.40515527000152973,0.32591397849462367,0.5948447299984703,0.6740860215053763
+0.7000000000000001,0.40561419611442556,0.32645161290322583,0.5943858038855745,0.6735483870967742
+0.7020000000000001,0.40622609759828665,0.3265591397849462,0.5937739024017133,0.6734408602150538
+0.704,0.4072969251950436,0.32784946236559137,0.5927030748049564,0.6721505376344086
+0.706,0.40737341288052625,0.3279569892473118,0.5926265871194738,0.6720430107526882
+0.708,0.4074499005660089,0.328494623655914,0.5925500994339912,0.671505376344086
+0.71,0.40844424047728317,0.3291397849462366,0.5915557595227168,0.6708602150537635
+0.712,0.4091326296466269,0.32978494623655913,0.5908673703533731,0.6702150537634408
+0.714,0.40928560501759215,0.32978494623655913,0.5907143949824079,0.6702150537634408
+0.716,0.4108918464127275,0.3313978494623656,0.5891081535872724,0.6686021505376344
+0.718,0.4120391616949671,0.33182795698924733,0.5879608383050329,0.6681720430107527
+0.72,0.41295701392075873,0.3325806451612903,0.5870429860792412,0.6674193548387097
+0.722,0.41341594003365456,0.3327956989247312,0.5865840599663454,0.6672043010752688
+0.724,0.4138748661465504,0.3331182795698925,0.5861251338534497,0.6668817204301075
+0.726,0.41441027994492885,0.33440860215053764,0.5855897200550712,0.6655913978494623
+0.728,0.41502218142878994,0.33483870967741935,0.58497781857121,0.6651612903225806
+0.73,0.41548110754168577,0.33505376344086024,0.5845188924583142,0.6649462365591398
+0.732,0.4156340829126511,0.33505376344086024,0.584365917087349,0.6649462365591398
+0.734,0.4199938809851614,0.3382795698924731,0.5800061190148387,0.6617204301075269
+0.736,0.4202233440416093,0.33838709677419354,0.5797766559583907,0.6616129032258065
+0.738,0.4203763194125746,0.33838709677419354,0.5796236805874254,0.6616129032258065
+0.74,0.4213706593238489,0.33903225806451615,0.5786293406761511,0.6609677419354838
+0.742,0.4221355361786752,0.3398924731182796,0.5778644638213248,0.6601075268817205
+0.744,0.42236499923512316,0.3401075268817204,0.5776350007648768,0.6598924731182796
+0.746,0.42297690071898425,0.3401075268817204,0.5770230992810157,0.6598924731182796
+0.748,0.42397124063025854,0.340752688172043,0.5760287593697415,0.6592473118279569
+0.75,0.4261128958237724,0.34236559139784944,0.5738871041762277,0.6576344086021506
+0.752,0.426189383509255,0.34236559139784944,0.573810616490745,0.6576344086021506
+0.754,0.42641884656570295,0.3432258064516129,0.573581153434297,0.6567741935483871
+0.756,0.42733669879149455,0.3446236559139785,0.5726633012085054,0.6553763440860215
+0.758,0.4277956249043904,0.3450537634408602,0.5722043750956096,0.6549462365591399
+0.76,0.42810157564632095,0.3450537634408602,0.5718984243536791,0.6549462365591399
+0.762,0.4292488909285605,0.3452688172043011,0.5707511090714394,0.654731182795699
+0.764,0.4292488909285605,0.3452688172043011,0.5707511090714394,0.654731182795699
+0.766,0.4302432308398348,0.34591397849462363,0.5697567691601653,0.6540860215053763
+0.768,0.43146703380755697,0.3467741935483871,0.568532966192443,0.6532258064516129
+0.77,0.4320024476059354,0.34763440860215056,0.5679975523940646,0.6523655913978494
+0.772,0.4326908367752792,0.34795698924731183,0.5673091632247208,0.6520430107526882
+0.774,0.4327673244607618,0.34795698924731183,0.5672326755392382,0.6520430107526882
+0.776,0.43284381214624446,0.34795698924731183,0.5671561878537555,0.6520430107526882
+0.778,0.4468410585895671,0.3606451612903226,0.553158941410433,0.6393548387096775
+0.78,0.44699403396053233,0.3606451612903226,0.5530059660394677,0.6393548387096775
+0.782,0.44714700933149765,0.3606451612903226,0.5528529906685024,0.6393548387096775
+0.784,0.4475294477589108,0.36075268817204303,0.5524705522410892,0.6392473118279569
+0.786,0.44806486155728925,0.3613978494623656,0.5519351384427107,0.6386021505376345
+0.788,0.44844729998470245,0.3620430107526882,0.5515527000152975,0.6379569892473118
+0.79,0.44921217683952885,0.3624731182795699,0.5507878231604711,0.63752688172043
+0.792,0.449594615266942,0.36301075268817207,0.550405384733058,0.6369892473118279
+0.794,0.44990056600887257,0.36301075268817207,0.5500994339911274,0.6369892473118279
+0.796,0.45028300443628577,0.3635483870967742,0.5497169955637142,0.6364516129032258
+0.798,0.45051246749273366,0.3635483870967742,0.5494875325072663,0.6364516129032258
+0.8,0.4516597827749732,0.36451612903225805,0.5483402172250268,0.635483870967742
+0.802,0.45188924583142115,0.3646236559139785,0.5481107541685788,0.6353763440860215
+0.804,0.45288358574269544,0.3662365591397849,0.5471164142573046,0.6337634408602151
+0.806,0.45357197491203916,0.3667741935483871,0.5464280250879608,0.633225806451613
+0.808,0.4536484625975218,0.3668817204301075,0.5463515374024782,0.6331182795698924
+0.81,0.45762582224261894,0.37150537634408604,0.5423741777573811,0.6284946236559139
+0.812,0.45770230992810157,0.37150537634408604,0.5422976900718984,0.6284946236559139
+0.8140000000000001,0.45808474835551477,0.3720430107526882,0.5419152516444852,0.6279569892473118
+0.8160000000000001,0.45892611289582375,0.3725806451612903,0.5410738871041763,0.6274193548387097
+0.8180000000000001,0.45907908826678906,0.37268817204301075,0.540920911733211,0.6273118279569893
+0.8200000000000001,0.4597674774361328,0.373010752688172,0.5402325225638672,0.6269892473118279
+0.8220000000000001,0.4602264035490286,0.3736559139784946,0.5397735964509713,0.6263440860215054
+0.8240000000000001,0.4612207434603029,0.37473118279569895,0.5387792565396972,0.625268817204301
+0.8260000000000001,0.4619091326296466,0.37548387096774194,0.5380908673703534,0.6245161290322581
+0.8280000000000001,0.462674009484473,0.3761290322580645,0.537325990515527,0.6238709677419355
+0.8300000000000001,0.46305644791188616,0.3761290322580645,0.5369435520881138,0.6238709677419355
+0.8320000000000001,0.4632859109683341,0.3761290322580645,0.536714089031666,0.6238709677419355
+0.834,0.4703992657182194,0.3846236559139785,0.5296007342817806,0.6153763440860215
+0.836,0.47062872877466727,0.3846236559139785,0.5293712712253327,0.6153763440860215
+0.838,0.47093467951659784,0.3846236559139785,0.5290653204834022,0.6153763440860215
+0.84,0.4718525317423895,0.3849462365591398,0.5281474682576105,0.6150537634408602
+0.842,0.47284687165366374,0.3851612903225806,0.5271531283463362,0.6148387096774194
+0.844,0.4734587731375249,0.38623655913978494,0.5265412268624752,0.6137634408602151
+0.846,0.47361174850849014,0.38623655913978494,0.5263882514915099,0.6137634408602151
+0.848,0.4748355514762123,0.3875268817204301,0.5251644485237876,0.61247311827957
+0.85,0.4755239406455561,0.38827956989247314,0.524476059354444,0.6117204301075269
+0.852,0.4778185712100352,0.39086021505376345,0.5221814287899649,0.6091397849462366
+0.854,0.478277497322931,0.39086021505376345,0.521722502677069,0.6091397849462366
+0.856,0.47911886186324004,0.3911827956989247,0.52088113813676,0.6088172043010753
+0.858,0.4802661771454796,0.3919354838709677,0.5197338228545204,0.6080645161290323
+0.86,0.4813370047422365,0.3931182795698925,0.5186629952577635,0.6068817204301076
+0.862,0.4817194431696497,0.3939784946236559,0.5182805568303503,0.606021505376344
+0.864,0.48225485696802817,0.39408602150537636,0.5177451430319718,0.6059139784946237
+0.866,0.482713783080924,0.39408602150537636,0.517286216919076,0.6059139784946237
+0.868,0.49173932996787517,0.39946236559139786,0.5082606700321248,0.6005376344086022
+0.87,0.4918158176533578,0.39946236559139786,0.5081841823466422,0.6005376344086022
+0.872,0.49204528070980574,0.39946236559139786,0.5079547192901943,0.6005376344086022
+0.874,0.4928866452501147,0.3998924731182796,0.5071133547498853,0.6001075268817204
+0.876,0.4934220590484932,0.40096774193548385,0.5065779409515068,0.5990322580645162
+0.878,0.49395747284687164,0.4010752688172043,0.5060425271531284,0.5989247311827957
+0.88,0.4952577635000765,0.4024731182795699,0.5047422364999234,0.5975268817204301
+0.882,0.49579317729845496,0.4024731182795699,0.504206822701545,0.5975268817204301
+0.884,0.4969404925806945,0.4043010752688172,0.5030595074193055,0.5956989247311828
+0.886,0.4977053694355209,0.4046236559139785,0.5022946305644791,0.5953763440860216
+0.888,0.4983937586048646,0.4050537634408602,0.5016062413951354,0.5949462365591398
+0.89,0.505277650298302,0.4147311827956989,0.49472234970169804,0.585268817204301
+0.892,0.5055836010402325,0.4147311827956989,0.4944163989597675,0.585268817204301
+0.894,0.5060425271531284,0.41483870967741937,0.49395747284687164,0.5851612903225807
+0.896,0.5077252562337463,0.41559139784946236,0.4922747437662537,0.5844086021505377
+0.898,0.5104788129111213,0.41731182795698923,0.4895211870888787,0.5826881720430108
+0.9,0.5114731528223956,0.41838709677419356,0.4885268471776044,0.5816129032258064
+0.902,0.5121615419917394,0.41838709677419356,0.48783845800826064,0.5816129032258064
+0.904,0.5124674927336699,0.41870967741935483,0.48753250726633013,0.5812903225806452
+0.906,0.5182040691448677,0.42376344086021506,0.4817959308551323,0.5762365591397849
+0.908,0.5187394829432461,0.42376344086021506,0.4812605170567539,0.5762365591397849
+0.91,0.5198867982254857,0.4243010752688172,0.48011320177451433,0.5756989247311828
+0.912,0.52088113813676,0.4260215053763441,0.47911886186324004,0.5739784946236559
+0.914,0.5213400642496558,0.42612903225806453,0.47865993575034416,0.5738709677419355
+0.916,0.5221814287899648,0.4266666666666667,0.47781857121003524,0.5733333333333333
+0.918,0.527000152975371,0.43086021505376343,0.47299984702462905,0.5691397849462365
+0.92,0.5272296160318188,0.4309677419354839,0.47277038396818116,0.5690322580645162
+0.922,0.5281474682576105,0.4324731182795699,0.4718525317423895,0.5675268817204301
+0.924,0.5283004436285758,0.4327956989247312,0.47169955637142424,0.5672043010752688
+0.926,0.5306715618785376,0.43548387096774194,0.4693284381214624,0.564516129032258
+0.928,0.5309010249349855,0.43623655913978493,0.4690989750650145,0.563763440860215
+0.93,0.5319718525317424,0.4375268817204301,0.4680281474682576,0.5624731182795699
+0.932,0.5322778032736729,0.4375268817204301,0.4677221967263271,0.5624731182795699
+0.934,0.5353373106929784,0.4417204301075269,0.4646626893070216,0.5582795698924732
+0.936,0.5357962368058743,0.44193548387096776,0.4642037631941257,0.5580645161290323
+0.9380000000000001,0.536714089031666,0.44344086021505374,0.46328591096833405,0.5565591397849463
+0.9400000000000001,0.5392381826525929,0.4486021505376344,0.46076181734740707,0.5513978494623656
+0.9420000000000001,0.5394676457090408,0.44870967741935486,0.4605323542909592,0.5512903225806451
+0.9440000000000001,0.5395441333945235,0.44870967741935486,0.46045586660547655,0.5512903225806451
+0.9460000000000001,0.5425271531283463,0.4513978494623656,0.4574728468716537,0.5486021505376344
+0.9480000000000001,0.542603640813829,0.4513978494623656,0.45739635918617105,0.5486021505376344
+0.9500000000000001,0.5437509560960685,0.4541935483870968,0.4562490439039315,0.5458064516129032
+0.9520000000000001,0.5443628575799296,0.45440860215053763,0.45563714242007036,0.5455913978494624
+0.9540000000000001,0.5458926112895823,0.45903225806451614,0.45410738871041767,0.5409677419354839
+0.9560000000000001,0.5482637295395442,0.4625806451612903,0.45173627046045584,0.5374193548387096
+0.9580000000000001,0.5491050940798532,0.4631182795698925,0.4508949059201468,0.5368817204301075
+0.96,0.551017286216919,0.4653763440860215,0.44898271378308097,0.5346236559139785
+0.962,0.5527000152975371,0.46623655913978496,0.4472999847024629,0.533763440860215
+0.964,0.5539238182652593,0.46806451612903227,0.44607618173474073,0.5319354838709678
+0.966,0.5558360104023252,0.46946236559139787,0.4441639895976748,0.5305376344086021
+0.968,0.5582836163377696,0.471505376344086,0.4417163836622304,0.528494623655914
+0.97,0.559889857732905,0.4720430107526882,0.440110142267095,0.5279569892473118
+0.972,0.5605017592167661,0.472258064516129,0.43949824078323385,0.527741935483871
+0.974,0.5609606853296619,0.47236559139784945,0.4390393146703381,0.5276344086021505
+0.976,0.5613431237570751,0.4726881720430108,0.43865687624292493,0.5273118279569893
+0.978,0.5617255621844883,0.4731182795698925,0.4382744378155117,0.5268817204301075
+0.98,0.5624904390393146,0.4743010752688172,0.4375095609606854,0.5256989247311827
+0.982,0.5627963897812452,0.4746236559139785,0.43720361021875476,0.5253763440860215
+0.984,0.5633318035796236,0.4750537634408602,0.43666819642037635,0.5249462365591397
+0.986,0.5634082912651063,0.47591397849462364,0.4365917087348937,0.5240860215053764
+0.988,0.5645556065473458,0.47655913978494624,0.4354443934526542,0.5234408602150538
+0.99,0.5649380449747591,0.4768817204301075,0.4350619550252409,0.5231182795698925
+0.992,0.5650910203457243,0.4770967741935484,0.43490897965427566,0.5229032258064517
+0.994,0.5663148233134465,0.47720430107526884,0.4336851766865535,0.5227956989247311
+0.996,0.5663913109989291,0.47774193548387095,0.43360868900107086,0.522258064516129
+0.998,0.5665442863698944,0.4786021505376344,0.4334557136301056,0.5213978494623657
+1.0,1.0,1.0,0.0,0.0

evaluation/query_family/subgroup/shared_basis_stability_analysis/support3_model_advantage.csv

@@ -0,0 +1,12 @@
+model_id,model_label,paired_panel_count,support3_diff,key_diff,query_diff,size_better_support3_share,ability_reading
+tvae,TVAE,44,0.03472692173824363,0.048947072520338114,0.0649191936769872,0.45454545454545453,Size-led scaffold preserver
+tabpfgen,TabPFGen,35,0.03387234329403705,0.030033846199508447,0.044555813417141966,0.5428571428571428,Size-led scaffold preserver
+tabddpm,TabDDPM,34,0.027154105591866824,0.04391652104647479,0.04900884774216029,0.5588235294117647,Moderate size-led scaffold preserver
+ctgan,CTGAN,43,0.026719379513754075,0.030908789607392725,0.049082595411640555,0.3953488372093023,Moderate size-led scaffold preserver
+bayesnet,BayesNet,45,0.022825237572080253,0.027431777023618487,0.046993088469907816,0.4444444444444444,Moderate size-led scaffold preserver
+tabdiff,TabDiff,16,0.009102120837056926,0.021879119137115887,0.05203784235631523,0.375,Near-balanced with mild size edge
+tabbyflow,TabbyFlow,30,0.0026109434970345347,0.03399328223872071,0.04966033047251702,0.4,Near-balanced with mild size edge
+realtabformer,RealTabFormer,39,0.001746449321126829,0.007762492602087272,0.03125283981632626,0.3333333333333333,Mixed or weak separation
+forestdiffusion,ForestDiffusion,25,-0.010863243153791037,-0.0024940425590155983,0.0025128727900994995,0.52,No clear scaffold edge
+arf,ARF,45,-0.01214084631782912,-0.001172867132097539,0.01602725432393037,0.4666666666666667,No clear scaffold edge
+tabsyn,TabSyn,42,-0.020398811610540867,0.03210385576340672,0.04357772628871713,0.3333333333333333,Key retention without stable size scaffold

evaluation/query_family/subgroup/shared_basis_stability_analysis/support3_model_advantage_bars.tex

@@ -0,0 +1,37 @@
+\documentclass[tikz,border=4pt]{standalone}
+\usepackage{pgfplots}
+\pgfplotsset{compat=1.18}
+\definecolor{modeltvae}{HTML}{4477AA}
+\definecolor{modeltabpfgen}{HTML}{009988}
+\definecolor{modeltabddpm}{HTML}{EE7733}
+\definecolor{modelctgan}{HTML}{EE6677}
+\definecolor{modelbayesnet}{HTML}{CCBB44}
+\definecolor{modeltabdiff}{HTML}{AA3377}
+\definecolor{modeltabbyflow}{HTML}{882255}
+\definecolor{modelrealtabformer}{HTML}{332288}
+\definecolor{modelforestdiffusion}{HTML}{228833}
+\definecolor{modelarf}{HTML}{777777}
+\definecolor{modeltabsyn}{HTML}{66CCEE}
+\begin{document}
+\begin{tikzpicture}
+\begin{axis}[
+width=17.0cm,
+height=8.7cm,
+ymin=-0.02, ymax=0.0900,
+xlabel={Model},
+ylabel={Mean panel support3(size) - support3(internal)},
+title={Model-level scaffold-preservation advantage},
+ymajorgrids,
+grid style={draw=gray!20},
+major grid style={draw=gray!30},
+axis line style={draw=black!70},
+tick style={draw=black!70},
+symbolic x coords={TVAE,TabPFGen,TabDDPM,CTGAN,BayesNet,TabDiff,TabbyFlow,RealTabFormer,ForestDiffusion,ARF,TabSyn},
+xtick=data,
+xticklabel style={rotate=40, anchor=east, font=\scriptsize},
+]
+\addplot[dashed, line width=1.0pt, color=black, opacity=0.85] coordinates {(TVAE,0) (TabSyn,0)};
+\addplot+[ybar, bar width=7.5pt, draw=modeltvae, fill=modeltvae] coordinates { (TVAE,0.034727) };\addplot+[ybar, bar width=7.5pt, draw=modeltabpfgen, fill=modeltabpfgen] coordinates { (TabPFGen,0.033872) };\addplot+[ybar, bar width=7.5pt, draw=modeltabddpm, fill=modeltabddpm] coordinates { (TabDDPM,0.027154) };\addplot+[ybar, bar width=7.5pt, draw=modelctgan, fill=modelctgan] coordinates { (CTGAN,0.026719) };\addplot+[ybar, bar width=7.5pt, draw=modelbayesnet, fill=modelbayesnet] coordinates { (BayesNet,0.022825) };\addplot+[ybar, bar width=7.5pt, draw=modeltabdiff, fill=modeltabdiff] coordinates { (TabDiff,0.009102) };\addplot+[ybar, bar width=7.5pt, draw=modeltabbyflow, fill=modeltabbyflow] coordinates { (TabbyFlow,0.002611) };\addplot+[ybar, bar width=7.5pt, draw=modelrealtabformer, fill=modelrealtabformer] coordinates { (RealTabFormer,0.001746) };\addplot+[ybar, bar width=7.5pt, draw=modelforestdiffusion, fill=modelforestdiffusion] coordinates { (ForestDiffusion,-0.010863) };\addplot+[ybar, bar width=7.5pt, draw=modelarf, fill=modelarf] coordinates { (ARF,-0.012141) };\addplot+[ybar, bar width=7.5pt, draw=modeltabsyn, fill=modeltabsyn] coordinates { (TabSyn,-0.020399) };
+\end{axis}
+\end{tikzpicture}
+\end{document}

evaluation/query_family/subgroup/shared_basis_stability_analysis/support3_paired_delta_hist.csv

@@ -0,0 +1,33 @@
+bin_left,bin_right,bin_center,density
+-0.4,-0.375,-0.3875,0.0
+-0.375,-0.35000000000000003,-0.36250000000000004,0.0
+-0.35000000000000003,-0.325,-0.3375,0.10050251256281398
+-0.325,-0.30000000000000004,-0.3125,0.10050251256281421
+-0.30000000000000004,-0.275,-0.28750000000000003,0.0
+-0.275,-0.25,-0.2625,0.0
+-0.25,-0.225,-0.2375,0.10050251256281409
+-0.225,-0.2,-0.21250000000000002,0.20100502512562818
+-0.2,-0.17500000000000002,-0.1875,0.3015075376884423
+-0.17500000000000002,-0.15000000000000002,-0.16250000000000003,0.20100502512562818
+-0.15000000000000002,-0.125,-0.1375,0.50251256281407
+-0.125,-0.09999999999999998,-0.11249999999999999,0.7035175879396979
+-0.09999999999999998,-0.07500000000000001,-0.0875,1.407035175879399
+-0.07500000000000001,-0.04999999999999999,-0.0625,1.9095477386934656
+-0.04999999999999999,-0.025000000000000022,-0.037500000000000006,4.321608040201011
+-0.025000000000000022,0.0,-0.012500000000000011,8.04020100502512
+0.0,0.025000000000000022,0.012500000000000011,10.753768844221096
+0.025000000000000022,0.04999999999999999,0.037500000000000006,2.512562814070355
+0.04999999999999999,0.07500000000000001,0.0625,2.1105527638190935
+0.07500000000000001,0.09999999999999998,0.0875,1.7085427135678415
+0.09999999999999998,0.125,0.11249999999999999,0.9045226130653258
+0.125,0.15000000000000002,0.1375,1.1055276381909538
+0.15000000000000002,0.17500000000000004,0.16250000000000003,0.9045226130653258
+0.17500000000000004,0.20000000000000007,0.18750000000000006,1.1055276381909538
+0.20000000000000007,0.22499999999999998,0.21250000000000002,0.20100502512562884
+0.22499999999999998,0.25,0.2375,0.20100502512562796
+0.25,0.275,0.2625,0.30150753768844196
+0.275,0.30000000000000004,0.28750000000000003,0.10050251256281398
+0.30000000000000004,0.32500000000000007,0.31250000000000006,0.0
+0.32500000000000007,0.35,0.3375,0.10050251256281442
+0.35,0.375,0.3625,0.10050251256281398
+0.375,0.4,0.3875,0.0

evaluation/query_family/subgroup/shared_basis_stability_analysis/support3_paired_delta_hist.tex

@@ -0,0 +1,44 @@
+\documentclass[tikz,border=4pt]{standalone}
+\usepackage{pgfplots}
+\pgfplotsset{compat=1.18}
+\definecolor{sizecolor}{HTML}{009988}
+\begin{document}
+\begin{tikzpicture}
+\begin{axis}[
+width=14.8cm,
+height=8.7cm,
+xmin=-0.40, xmax=0.40,
+ymin=0.0, ymax=11.8291,
+xlabel={Panel-level support3(size) - support3(internal)},
+ylabel={Relative density},
+title={Matched-panel subgroup support advantage},
+ymajorgrids,
+grid style={draw=gray!20},
+major grid style={draw=gray!30},
+axis line style={draw=black!70},
+tick style={draw=black!70},
+]
+\addplot[ybar interval, mark=none, draw=white, fill=sizecolor, fill opacity=0.30]
+table[x=bin_left, y=density, col sep=comma]{support3_paired_delta_hist.csv};
+\addplot[dashed, line width=1.0pt, color=black, opacity=0.85]
+coordinates {(0,0) (0,11.8291)};
+\addplot[line width=1.6pt, color=sizecolor]
+coordinates {(0.011131,0) (0.011131,11.8291)};
+\node[
+anchor=north west,
+draw=gray!55,
+fill=white,
+rounded corners=2pt,
+fill opacity=0.96,
+text opacity=1,
+align=left,
+font=\small
+] at (axis description cs:0.015,0.97) {
+Panels: 398\\
+Mean delta: 0.011\\
+Share > 0: 0.437\\
+Share >= 0.10: 0.126
+};
+\end{axis}
+\end{tikzpicture}
+\end{document}

evaluation/query_family/subgroup/shared_basis_stability_analysis/support3_paired_panel_table.csv

@@ -0,0 +1,399 @@
+dataset_prefix,dataset_id,model_id,model_label,support3_score__internal_profile_stability,support3_score__subgroup_size_stability,key_set_score__internal_profile_stability,key_set_score__subgroup_size_stability,query_score__internal_profile_stability,query_score__subgroup_size_stability,support3_delta_size_minus_internal,key_delta_size_minus_internal,query_delta_size_minus_internal
+c,c10,arf,ARF,0.6599999999999999,0.4155432037703214,0.0,0.0,0.0,0.0,-0.24445679622967853,0.0,0.0
+c,c10,bayesnet,BayesNet,0.3333333333333333,0.4155428319332523,0.0,0.0,0.0,0.0,0.082209498599919,0.0,0.0
+c,c10,ctgan,CTGAN,0.9552380952380952,0.9838217338217338,0.9328571428571429,0.9757326007326008,0.9328571,0.9757325384615385,0.02858363858363855,0.04287545787545788,0.04287543846153852
+c,c10,forestdiffusion,ForestDiffusion,0.4438461538461539,0.4155429239185576,0.0,0.0,0.0,0.0,-0.028303229927596263,0.0,0.0
+c,c10,tabbyflow,TabbyFlow,0.6866666666666666,0.5284944150227905,0.08,0.0,0.08,0.0,-0.15817225164387616,-0.08,-0.08
+c,c10,tabddpm,TabDDPM,0.6633333333333333,0.6538461538461539,0.0,0.0,0.0,0.0,-0.009487179487179476,0.0,0.0
+c,c10,tabpfgen,TabPFGen,0.3333333333333333,0.416414453190324,0.0,0.0,0.0,0.0,0.08308111985699068,0.0,0.0
+c,c10,tabsyn,TabSyn,0.6533333333333333,0.531487480297553,0.0,0.0,0.0,0.0,-0.12184585303578033,0.0,0.0
+c,c10,tvae,TVAE,0.934908424908425,0.9387386118155349,0.8954945054945055,0.9002888137503522,0.8954945000000001,0.9002888461538461,0.0038301869071099004,0.004794308255846702,0.004794346153845996
+c,c11,arf,ARF,0.9911111111111112,0.9730639730639732,0.9866666666666667,0.9595959595959597,0.98666668,0.959596,-0.01804713804713798,-0.027070707070707023,-0.027070680000000014
+c,c11,bayesnet,BayesNet,1.0,0.9932659932659933,1.0,0.9898989898989898,1.0,0.9898989999999999,-0.006734006734006703,-0.010101010101010166,-0.010101000000000138
+c,c11,ctgan,CTGAN,0.9911111111111112,0.9730639730639732,0.9866666666666667,0.9595959595959597,0.98666668,0.959596,-0.01804713804713798,-0.027070707070707023,-0.027070680000000014
+c,c11,forestdiffusion,ForestDiffusion,0.48,0.48148148148148145,0.0,0.0,0.0,0.0,0.0014814814814814725,0.0,0.0
+c,c11,realtabformer,RealTabFormer,1.0,0.9977553310886643,1.0,0.9966329966329965,1.0,0.996633,-0.002244668911335679,-0.0033670033670034627,-0.003367000000000009
+c,c11,tabbyflow,TabbyFlow,0.48,0.48484848484848486,0.0,0.0,0.0,0.0,0.00484848484848488,0.0,0.0
+c,c11,tabddpm,TabDDPM,0.48,0.48484848484848486,0.0,0.0,0.0,0.0,0.00484848484848488,0.0,0.0
+c,c11,tabsyn,TabSyn,0.4755555555555555,0.46037346491891945,0.0,0.0,0.0,0.0,-0.015182090636636059,0.0,0.0
+c,c11,tvae,TVAE,1.0,0.978675645342312,1.0,0.968013468013468,1.0,0.9680134848484849,-0.021324354657688005,-0.03198653198653201,-0.03198651515151507
+c,c12,arf,ARF,0.9824561403508772,0.9871794871794872,0.9736842105263158,0.9807692307692307,0.9736842105263158,0.9807692307692307,0.004723346828609976,0.007085020242914908,0.007085020242914908
+c,c12,bayesnet,BayesNet,1.0,0.9871794871794872,1.0,0.9807692307692307,1.0,0.9807692307692307,-0.012820512820512775,-0.019230769230769273,-0.019230769230769273
+c,c12,ctgan,CTGAN,0.9912280701754386,0.9743589743589743,0.9868421052631579,0.9615384615384616,0.9868421052631579,0.9615384615384616,-0.01686909581646423,-0.02530364372469629,-0.02530364372469629
+c,c12,realtabformer,RealTabFormer,0.9912280701754386,1.0,0.9868421052631579,1.0,0.9868421052631579,1.0,0.00877192982456143,0.013157894736842146,0.013157894736842146
+c,c12,tabpfgen,TabPFGen,0.9649122807017543,1.0,0.9473684210526315,1.0,0.9473684210526315,1.0,0.03508771929824572,0.052631578947368474,0.052631578947368474
+c,c12,tvae,TVAE,0.6842105263157895,0.6538461538461539,0.5263157894736842,0.4807692307692308,0.5263157894736842,0.4807692307692308,-0.030364372469635637,-0.0455465587044534,-0.0455465587044534
+c,c14,arf,ARF,0.8256302521008403,0.6451107160243154,0.6764705882352942,0.38159340659340657,0.6764705882352942,0.38159342307692307,-0.18051953607652493,-0.2948771816418876,-0.2948771651583711
+c,c14,bayesnet,BayesNet,0.8859477124183007,0.910998955795336,0.8196078431372549,0.8096153846153846,0.8196078529411764,0.8096153846153846,0.025051243377035237,-0.009992458521870251,-0.009992468325791815
+c,c14,ctgan,CTGAN,0.8855042016806722,0.8644476438594085,0.8245798319327731,0.7751028383381324,0.8245797941176471,0.7751028461538462,-0.0210565578212637,-0.04947699359464064,-0.04947694796380098
+c,c14,realtabformer,RealTabFormer,0.9986928104575163,0.9772477522477523,0.996078431372549,0.9477688977688978,0.9960784411764706,0.9477689230769231,-0.021445058209764056,-0.0483095336036512,-0.04830951809954753
+c,c14,tabsyn,TabSyn,0.6094771241830065,0.5052790346907994,0.0,0.15384615384615385,0.0,0.15384615384615385,-0.1041980894922071,0.15384615384615385,0.15384615384615385
+c,c14,tvae,TVAE,0.6055645335057099,0.6677267177267178,0.4046703296703297,0.5015900765900766,0.40467038235294117,0.5015901153846154,0.06216218422100783,0.09691974691974692,0.09691973303167423
+c,c15,arf,ARF,0.9061240903346167,0.7052781799387063,0.8520164046479836,0.5196886002806624,0.8520164545454545,0.5196885769230769,-0.20084591039591038,-0.3323278043673211,-0.3323278776223776
+c,c15,bayesnet,BayesNet,0.8120176120176121,0.6786934292245654,0.6939393939393939,0.4258928571428572,0.6939394545454545,0.4258929230769231,-0.1333241827930467,-0.26804653679653667,-0.26804653146853136
+c,c15,tabsyn,TabSyn,0.6333968176073439,0.4468587169293745,0.03383458646616541,0.049404134773279096,0.033834636363636365,0.04940411538461539,-0.1865381006779694,0.015569548307113684,0.015569479020979021
+c,c15,tvae,TVAE,0.6528725903725904,0.4807132872901079,0.4602272727272727,0.18766219170097675,0.4602271818181818,0.18766215384615384,-0.17215930308248245,-0.27256508102629595,-0.272565027972028
+c,c16,arf,ARF,0.8862214465373444,0.918460688866233,0.7769365977920099,0.8004807692307693,0.7769366153846153,0.8004807692307693,0.03223924232888864,0.02354417143875942,0.023544153846153937
+c,c16,bayesnet,BayesNet,0.7512421633120496,0.7540718147927247,0.5665298558097055,0.5451717619986851,0.5665299038461539,0.5451717307692308,0.002829651480675044,-0.021358093811020384,-0.021358173076923115
+c,c16,ctgan,CTGAN,0.8019950690425852,0.7641570446508029,0.6600592239717626,0.5980769230769231,0.6600592115384616,0.5980769615384616,-0.0378380243917823,-0.06198230089483958,-0.061982250000000016
+c,c16,realtabformer,RealTabFormer,0.8610322057981716,0.861035595995683,0.7154773358106621,0.7091422909456291,0.7154773846153847,0.7091422692307692,3.3901975113881377e-06,-0.006335044865033002,-0.006335115384615508
+c,c16,tabsyn,TabSyn,0.5506382929409734,0.48450835473307385,0.0,0.0,0.0,0.0,-0.06612993820789959,0.0,0.0
+c,c16,tvae,TVAE,0.5840938042869094,0.5857831227472133,0.33442995389026225,0.33644345187128244,0.33442990384615384,0.3364434230769231,0.0016893184603039568,0.0020134979810201847,0.002013519230769256
+c,c17,arf,ARF,0.9017516434183102,0.8223942470732276,0.7121993746993747,0.6511876307744883,0.7121994444444444,0.6511876551724138,-0.07935739634508254,-0.0610117439248864,-0.06101178927203055
+c,c17,bayesnet,BayesNet,0.9057667316439245,0.8233419666359909,0.7242446393762183,0.65388702059164,0.7242446666666668,0.6538870344827586,-0.08242476500793361,-0.07035761878457836,-0.07035763218390811
+c,c17,ctgan,CTGAN,0.9384837855770088,0.9350486229172105,0.8154513567310265,0.8423155955765308,0.8154513333333333,0.8423155862068965,-0.003435162659798374,0.026864238845504262,0.0268642528735632
+c,c17,realtabformer,RealTabFormer,0.8791876025209359,0.8366856505986048,0.7013092463092464,0.6873284002123504,0.7013092777777777,0.6873283793103449,-0.04250195192233108,-0.013980846096895982,-0.01398089846743289
+c,c17,tabpfgen,TabPFGen,0.8919951709273959,0.8198189406099676,0.6829299572266324,0.6447729911151779,0.68293,0.6447729999999999,-0.07217623031742826,-0.03815696611145447,-0.03815700000000011
+c,c17,tabsyn,TabSyn,0.6643518518518517,0.4703846576976359,0.0,0.0,0.0,0.0,-0.19396719415421587,0.0,0.0
+c,c17,tvae,TVAE,0.9018804241026463,0.8730880553909004,0.7535401327067994,0.7583263889206753,0.7535401111111111,0.7583264137931034,-0.028792368711745953,0.00478625621387585,0.004786302681992316
+c,c18,arf,ARF,0.7476247118271779,0.7539227822330198,0.45837955957280485,0.5620930197828807,0.366703675,0.562093,0.006298070405841849,0.10371346021007588,0.19538932499999995
+c,c18,bayesnet,BayesNet,0.6889623114034521,0.7524087516309592,0.3876768647659119,0.49061069126353823,0.310141475,0.49061073913043485,0.06344644022750712,0.10293382649762634,0.18046926413043485
+c,c18,realtabformer,RealTabFormer,0.8273187879861233,0.8399302574806287,0.7082945490104974,0.7345330673803633,0.566635675,0.7345330869565218,0.012611469494505312,0.026238518369865904,0.16789741195652175
+c,c18,tabsyn,TabSyn,0.5721335383274612,0.495492780778945,0.048006221719457016,0.049858666989978664,0.03840495,0.049858695652173914,-0.07664075754851618,0.001852445270521648,0.011453745652173913
+c,c19,arf,ARF,0.999544315333789,1.0,0.9993164730006835,1.0,0.9993164675324675,1.0,0.00045568466621104253,0.0006835269993165083,0.0006835324675324994
+c,c19,bayesnet,BayesNet,0.7771793199966298,0.809570325275933,0.5995670995670995,0.7142857142857143,0.5995671038961039,0.7142857142857143,0.032391005279303164,0.11471861471861478,0.1147186103896104
+c,c19,ctgan,CTGAN,0.7506496422714106,0.8385854341736695,0.5981796024194802,0.7478991596638656,0.5981795974025974,0.7478991428571428,0.08793579190225886,0.14971955724438535,0.14971954545454536
+c,c19,realtabformer,RealTabFormer,0.6717191258564745,0.6904761904761905,0.5033480709177304,0.5357142857142857,0.5033480649350649,0.5357142857142857,0.018757064619715935,0.032366214796555326,0.0323662207792208
+c,c19,tabpfgen,TabPFGen,0.38944765059348907,0.33339486944309066,0.0,0.0,0.0,0.0,-0.05605278115039841,0.0,0.0
+c,c19,tabsyn,TabSyn,0.5696743640593908,0.551693404634581,0.0,0.0,0.0,0.0,-0.017980959424809706,0.0,0.0
+c,c19,tvae,TVAE,0.7587350180385547,0.843915343915344,0.6031251341116156,0.746031746031746,0.6031251168831169,0.7460317142857144,0.08518032587678925,0.14290661192013043,0.14290659740259748
+c,c2,arf,ARF,1.0,0.9948839039748131,1.0,0.9901613537977174,1.0,0.9901613636363636,-0.0051160960251869225,-0.009838646202282586,-0.009838636363636355
+c,c2,bayesnet,BayesNet,1.0,0.9985569985569984,1.0,0.9956709956709956,1.0,0.995671,-0.0014430014430015792,-0.0043290043290044045,-0.004329000000000027
+c,c2,ctgan,CTGAN,1.0,0.9919498010407102,1.0,0.9813590449954086,1.0,0.9813590303030303,-0.008050198959289823,-0.018640955004591397,-0.01864096969696971
+c,c2,forestdiffusion,ForestDiffusion,0.4622222222222222,0.4729430706866159,0.0,0.0,0.0,0.0,0.010720848464393695,0.0,0.0
+c,c2,realtabformer,RealTabFormer,1.0,0.9985569985569984,1.0,0.9956709956709956,1.0,0.995671,-0.0014430014430015792,-0.0043290043290044045,-0.004329000000000027
+c,c2,tabbyflow,TabbyFlow,0.5182222222222223,0.4902356902356902,0.07466666666666667,0.01616161616161616,0.07466664,0.016161606060606064,-0.02798653198653206,-0.05850505050505051,-0.05850503393939394
+c,c2,tabddpm,TabDDPM,0.48,0.483011937557392,0.0,0.0,0.0,0.0,0.0030119375573920415,0.0,0.0
+c,c2,tabdiff,TabDiff,0.5182222222222223,0.4883991429445974,0.07466666666666667,0.01616161616161616,0.07466664,0.016161606060606064,-0.029823079277624842,-0.05850505050505051,-0.05850503393939394
+c,c2,tabpfgen,TabPFGen,0.94,0.8866873866873867,0.91,0.8300310800310801,0.91,0.8300310606060606,-0.05331261331261328,-0.07996891996891997,-0.07996893939393945
+c,c2,tabsyn,TabSyn,0.5093333333333333,0.48686868686868684,0.07466666666666667,0.01616161616161616,0.07466664,0.016161606060606064,-0.022464646464646465,-0.05850505050505051,-0.05850503393939394
+c,c2,tvae,TVAE,1.0,0.9932004022913113,1.0,0.9851108487472124,1.0,0.9851108484848484,-0.006799597708688654,-0.014889151252787558,-0.014889151515151577
+c,c20,arf,ARF,0.452966218145544,0.5468291887562875,0.08,0.15151515151515152,0.06060606060606061,0.15151515151515152,0.0938629706107435,0.07151515151515152,0.09090909090909091
+c,c20,bayesnet,BayesNet,0.9871604938271605,0.9776454067522649,0.9614814814814815,0.9329362202567943,0.7283950606060606,0.9329362424242423,-0.009515087074895678,-0.028545261224687146,0.2045411818181817
+c,c20,ctgan,CTGAN,0.9947015250544662,0.9566034909172164,0.9849934640522876,0.8698104727516492,0.7462071818181818,0.8698105151515152,-0.03809803413724977,-0.11518299130063836,0.1236033333333334
+c,c20,forestdiffusion,ForestDiffusion,0.3425677378787262,0.4438281017766766,0.0,0.0,0.0,0.0,0.1012603638979504,0.0,0.0
+c,c20,realtabformer,RealTabFormer,0.9982222222222222,0.9769666360575452,0.9946666666666667,0.9308999081726355,0.7535353636363636,0.9308999090909091,-0.021255586164677043,-0.06376675849403124,0.1773645454545455
+c,c20,tabbyflow,TabbyFlow,0.4412107749357231,0.4825577612493375,0.0,0.0,0.0,0.0,0.04134698631361444,0.0,0.0
+c,c20,tabddpm,TabDDPM,0.6666666666666665,0.6363636363636364,0.0,0.0,0.0,0.0,-0.030303030303030165,0.0,0.0
+c,c20,tabpfgen,TabPFGen,0.38674085770032285,0.5468135519006909,0.08,0.15151515151515152,0.06060606060606061,0.15151515151515152,0.16007269420036807,0.07151515151515152,0.09090909090909091
+c,c20,tabsyn,TabSyn,0.5520423631224091,0.5824547823431315,0.0,0.0,0.0,0.0,0.030412419220722464,0.0,0.0
+c,c20,tvae,TVAE,0.9803430791532931,0.9603487327945532,0.9548624448168961,0.8810461983836596,0.7233806060606061,0.8810461515151515,-0.019994346358739867,-0.07381624643323648,0.15766554545454547
+c,c3,arf,ARF,1.0,1.0,1.0,1.0,1.0,1.0,0.0,0.0,0.0
+c,c3,bayesnet,BayesNet,1.0,1.0,1.0,1.0,1.0,1.0,0.0,0.0,0.0
+c,c3,ctgan,CTGAN,1.0,1.0,1.0,1.0,1.0,1.0,0.0,0.0,0.0
+c,c3,realtabformer,RealTabFormer,1.0,1.0,1.0,1.0,1.0,1.0,0.0,0.0,0.0
+c,c3,tabbyflow,TabbyFlow,1.0,1.0,1.0,1.0,1.0,1.0,0.0,0.0,0.0
+c,c3,tabddpm,TabDDPM,1.0,1.0,1.0,1.0,1.0,1.0,0.0,0.0,0.0
+c,c3,tabdiff,TabDiff,1.0,1.0,1.0,1.0,1.0,1.0,0.0,0.0,0.0
+c,c3,tabpfgen,TabPFGen,1.0,1.0,1.0,1.0,1.0,1.0,0.0,0.0,0.0
+c,c3,tabsyn,TabSyn,0.47619047619047616,0.43478260869565216,0.0,0.0,0.0,0.0,-0.041407867494824,0.0,0.0
+c,c3,tvae,TVAE,1.0,1.0,1.0,1.0,1.0,1.0,0.0,0.0,0.0
+c,c4,arf,ARF,0.9866666666666667,0.994949494949495,0.98,0.9924242424242424,0.98,0.9924242424242424,0.008282828282828336,0.012424242424242449,0.012424242424242449
+c,c4,bayesnet,BayesNet,1.0,0.9797979797979799,1.0,0.9621212121212122,1.0,0.9621212121212122,-0.02020202020202011,-0.037878787878787845,-0.037878787878787845
+c,c4,ctgan,CTGAN,1.0,0.9797979797979799,1.0,0.9696969696969697,1.0,0.9696969696969697,-0.02020202020202011,-0.030303030303030276,-0.030303030303030276
+c,c4,forestdiffusion,ForestDiffusion,0.96,0.962962962962963,0.94,0.9444444444444445,0.94,0.9444444545454546,0.002962962962963056,0.004444444444444584,0.0044444545454546125
+c,c4,realtabformer,RealTabFormer,1.0,0.994949494949495,1.0,0.9924242424242424,1.0,0.9924242424242424,-0.005050505050504972,-0.007575757575757569,-0.007575757575757569
+c,c4,tabbyflow,TabbyFlow,1.0,0.9797979797979799,1.0,0.9696969696969697,1.0,0.9696969696969697,-0.02020202020202011,-0.030303030303030276,-0.030303030303030276
+c,c4,tabddpm,TabDDPM,0.9866666666666667,0.9932659932659933,0.98,0.9898989898989898,0.98,0.9898989999999999,0.006599326599326605,0.009898989898989852,0.00989899999999988
+c,c4,tabdiff,TabDiff,0.9866666666666667,0.9848484848484849,0.98,0.9772727272727273,0.98,0.9772727272727273,-0.0018181818181818299,-0.0027272727272726893,-0.0027272727272726893
+c,c4,tabpfgen,TabPFGen,0.9866666666666667,0.9932659932659933,0.98,0.9898989898989898,0.98,0.9898989999999999,0.006599326599326605,0.009898989898989852,0.00989899999999988
+c,c4,tabsyn,TabSyn,0.96,0.9730639730639732,0.94,0.9595959595959597,0.94,0.9595959696969697,0.013063973063973222,0.019595959595959722,0.01959596969696975
+c,c4,tvae,TVAE,0.9866666666666667,0.9831649831649831,0.98,0.9747474747474747,0.98,0.9747474848484847,-0.0035016835016835612,-0.005252525252525286,-0.005252515151515258
+c,c5,arf,ARF,0.9888888888888888,0.969202907081695,0.9833333333333333,0.9468113536295355,0.98333332,0.9468113636363636,-0.01968598180719383,-0.03652197970379778,-0.03652195636363631
+c,c5,bayesnet,BayesNet,0.9664444444444444,0.9136154205532195,0.9426666666666667,0.8491566381518535,0.9426666400000001,0.8491566666666667,-0.0528290238912249,-0.09351002851481316,-0.09350997333333344
+c,c5,ctgan,CTGAN,0.9648888888888889,0.9430007142128354,0.9473333333333334,0.9056626874808693,0.94733332,0.905662696969697,-0.02188817467605353,-0.04167064585246405,-0.041670623030303
+c,c5,realtabformer,RealTabFormer,0.9955555555555555,0.9873715510079147,0.9933333333333333,0.9703791158336613,0.99333332,0.9703791212121212,-0.008184004547640789,-0.02295421749967197,-0.022954198787878743
+c,c5,tabbyflow,TabbyFlow,1.0,0.9890244435698982,1.0,0.9728584546766365,1.0,0.9728584545454545,-0.01097555643010184,-0.027141545323363547,-0.0271415454545455
+c,c5,tabddpm,TabDDPM,0.9933333333333333,0.9587287011529436,0.99,0.9279920416284053,0.99,0.9279920606060605,-0.03460463218038967,-0.062007958371594674,-0.06200793939393945
+c,c5,tabdiff,TabDiff,0.9715555555555555,0.935873890419345,0.9573333333333334,0.8967401285583104,0.95733332,0.8967401515151515,-0.035681665136210516,-0.060593204775022946,-0.060593168484848525
+c,c5,tabpfgen,TabPFGen,0.9834074074074074,0.9825710204498083,0.975111111111111,0.9651985220167039,0.97511112,0.9651985151515152,-0.0008363869575991334,-0.009912589094407176,-0.009912604848484774
+c,c5,tabsyn,TabSyn,0.9888888888888888,0.945732477518422,0.9833333333333333,0.8913211367756821,0.98333332,0.8913211818181819,-0.043156411370466774,-0.09201219655765114,-0.09201213818181808
+c,c5,tvae,TVAE,0.9466666666666668,0.9219952045709622,0.92,0.8683807774716866,0.92,0.8683808181818181,-0.024671462095704588,-0.05161922252831341,-0.051619181818181925
+c,c6,arf,ARF,0.9777777777777779,0.9418382383899625,0.9551587301587302,0.8950959959580649,0.9551587619047618,0.8950960344827587,-0.035939539387815356,-0.060062734200665324,-0.06006272742200314
+c,c6,bayesnet,BayesNet,0.9777777777777779,0.9728417176693039,0.9551587301587302,0.942170473204956,0.9551587619047618,0.9421704827586207,-0.004936060108473916,-0.012988256953774258,-0.01298827914614109
+c,c6,ctgan,CTGAN,0.8612407926133416,0.8096752582553799,0.7636165577342047,0.684378180980615,0.7636165714285715,0.6843782068965517,-0.05156553435796174,-0.07923837675358969,-0.07923836453201982
+c,c6,realtabformer,RealTabFormer,1.0,0.9896930655551346,1.0,0.9748263231021852,1.0,0.9748263103448276,-0.010306934444865412,-0.0251736768978148,-0.025173689655172415
+c,c6,tabbyflow,TabbyFlow,0.9644657863145257,0.9423145181765872,0.9299619847939176,0.8770256563360012,0.9299620000000001,0.8770257586206897,-0.02215126813793855,-0.052936328457916404,-0.05293624137931041
+c,c6,tabdiff,TabDiff,0.9630763416477702,0.9242719642313963,0.9348639455782313,0.8645758568679461,0.934864,0.8645758965517241,-0.03880437741637388,-0.07028808871028513,-0.07028810344827596
+c,c6,tabpfgen,TabPFGen,0.9957671957671957,0.9830407140751969,0.9873015873015873,0.9491221422255904,0.987301619047619,0.9491222068965517,-0.012726481691998859,-0.03817944507599691,-0.03817941215106735
+c,c6,tabsyn,TabSyn,0.9732804232804233,0.9591890867752937,0.9484126984126985,0.9264178350385247,0.9484127142857143,0.9264178620689655,-0.014091336505129615,-0.02199486337417378,-0.021994852216748795
+c,c6,tvae,TVAE,0.8834378195722734,0.829116544938046,0.8090436174469788,0.7198470224839394,0.8090436666666666,0.7198470689655173,-0.05432127463422742,-0.08919659496303944,-0.0891965977011493
+c,c7,arf,ARF,0.968,0.9966094511549056,0.9520000000000001,0.9921593557957193,0.9520000000000001,0.9921593636363636,0.028609451154905652,0.040159355795719276,0.04015936363636352
+c,c7,bayesnet,BayesNet,1.0,0.9981634527089072,1.0,0.9944903581267218,1.0,0.9944903636363637,-0.0018365472910928382,-0.0055096418732781816,-0.005509636363636328
+c,c7,ctgan,CTGAN,0.9946666666666667,0.9911495911495912,0.992,0.9816738816738816,0.992,0.9816738787878787,-0.003517075517075474,-0.010326118326118405,-0.010326121212121286
+c,c7,forestdiffusion,ForestDiffusion,0.9589473684210525,0.9831083948731008,0.9284210526315789,0.9746625923096511,0.92842104,0.9746625757575758,0.024161026452048207,0.04624153967807221,0.046241535757575725
+c,c7,realtabformer,RealTabFormer,1.0,0.9981634527089072,1.0,0.9944903581267218,1.0,0.9944903636363637,-0.0018365472910928382,-0.0055096418732781816,-0.005509636363636328
+c,c7,tabbyflow,TabbyFlow,0.9546666666666667,0.9214008214008214,0.932,0.8754578754578755,0.932,0.8754578484848484,-0.033265845265845306,-0.056542124542124594,-0.05654215151515163
+c,c7,tabddpm,TabDDPM,1.0,0.9967204512659059,1.0,0.9901613537977174,1.0,0.9901613636363636,-0.0032795487340940843,-0.009838646202282586,-0.009838636363636355
+c,c7,tabdiff,TabDiff,0.96,0.9214008214008214,0.94,0.8754578754578755,0.94,0.8754578484848484,-0.03859917859917861,-0.06454212454212449,-0.06454215151515152
+c,c7,tabpfgen,TabPFGen,0.9640000000000001,0.9189800098891008,0.946,0.8764038991311718,0.946,0.8764038787878787,-0.0450199901108993,-0.06959610086882817,-0.06959612121212122
+c,c7,tabsyn,TabSyn,0.8746666666666667,0.9397047397047397,0.812,0.9032634032634033,0.812,0.9032633939393939,0.06503807303807296,0.09126340326340321,0.09126339393939387
+c,c7,tvae,TVAE,0.956,0.9887334887334889,0.934,0.9831002331002331,0.934,0.9831002424242423,0.032733488733488914,0.04910023310023304,0.04910024242424227
+c,c8,arf,ARF,1.0,1.0,1.0,1.0,1.0,1.0,0.0,0.0,0.0
+c,c8,bayesnet,BayesNet,1.0,1.0,1.0,1.0,1.0,1.0,0.0,0.0,0.0
+c,c8,ctgan,CTGAN,1.0,1.0,1.0,1.0,1.0,1.0,0.0,0.0,0.0
+c,c8,forestdiffusion,ForestDiffusion,0.9666666666666666,0.9358974358974359,0.9,0.8846153846153846,0.9,0.8846153846153846,-0.03076923076923066,-0.015384615384615441,-0.015384615384615441
+c,c8,realtabformer,RealTabFormer,1.0,1.0,1.0,1.0,1.0,1.0,0.0,0.0,0.0
+c,c8,tabbyflow,TabbyFlow,0.6888888888888889,0.6304878696183044,0.1,0.08974358974358974,0.1,0.08974357692307693,-0.0584010192705845,-0.010256410256410262,-0.01025642307692308
+c,c8,tabddpm,TabDDPM,0.6555555555555556,0.6452991452991453,0.0,0.038461538461538464,0.0,0.038461538461538464,-0.01025641025641022,0.038461538461538464,0.038461538461538464
+c,c8,tabdiff,TabDiff,0.7,0.6249068595222441,0.1,0.08974358974358974,0.1,0.08974357692307693,-0.07509314047775584,-0.010256410256410262,-0.01025642307692308
+c,c8,tabpfgen,TabPFGen,1.0,1.0,1.0,1.0,1.0,1.0,0.0,0.0,0.0
+c,c8,tabsyn,TabSyn,0.7666666666666666,0.7692307692307693,0.3,0.3076923076923077,0.3,0.3076923076923077,0.002564102564102666,0.0076923076923077205,0.0076923076923077205
+c,c8,tvae,TVAE,0.9777777777777779,0.9743589743589743,0.9666666666666666,0.9615384615384616,0.9666667,0.9615385000000001,-0.0034188034188035177,-0.005128205128204999,-0.005128199999999916
+c,c9,arf,ARF,0.8339143542922696,0.867752689171613,0.501743062876809,0.604385889366325,0.50174296,0.6043858787878787,0.0338383348793434,0.10264282648951595,0.10264291878787879
+c,c9,bayesnet,BayesNet,0.7793514651326686,0.6782296970037148,0.33805439539800597,0.20682348299290748,0.33805436,0.20682351515151515,-0.10112176812895379,-0.1312309124050985,-0.13123084484848485
+c,c9,ctgan,CTGAN,0.7745679012345679,0.6729922416772064,0.32370370370370366,0.19159195475008456,0.32370372000000003,0.1915919696969697,-0.10157565955736148,-0.1321117489536191,-0.13211175030303032
+c,c9,forestdiffusion,ForestDiffusion,0.6294918418079364,0.6985624301836958,0.048475525423809174,0.22305357972763049,0.04847552,0.22305360606060606,0.06907058837575941,0.1745780543038213,0.17457808606060607
+c,c9,realtabformer,RealTabFormer,0.9637607480239059,0.9647518763534726,0.8912822440717177,0.8984036832503759,0.89128228,0.8984037272727273,0.000991128329566715,0.007121439178658151,0.007121447272727344
+c,c9,tabbyflow,TabbyFlow,0.8709260458839406,0.9306427571499291,0.6127781376518219,0.7972120554621182,0.61277812,0.7972120909090908,0.05971671126598843,0.18443391781029628,0.18443397090909086
+c,c9,tabddpm,TabDDPM,0.668216631674513,0.6780305294359956,0.16464989502353883,0.23359279184776652,0.16464991999999998,0.23359275757575756,0.00981389776148267,0.06894289682422769,0.06894283757575759
+c,c9,tabdiff,TabDiff,0.8756271540108933,0.932816529246712,0.6268814620326801,0.8036001440257446,0.62688156,0.8036001515151515,0.057189375235818685,0.1767186819930645,0.17671859151515157
+c,c9,tabpfgen,TabPFGen,0.8245825239088397,0.8426344108830083,0.4737475717265191,0.5368468241496116,0.47374760000000005,0.5368468484848484,0.018051886974168596,0.0630992524230925,0.06309924848484838
+c,c9,tabsyn,TabSyn,0.8720438787906281,0.9284868784519033,0.6161316363718841,0.7905405263837855,0.6161316800000001,0.7905405757575757,0.056442999661275195,0.1744088900119014,0.17440889575757568
+c,c9,tvae,TVAE,0.749405399218268,0.6638160074364088,0.288216197654804,0.1786058321921539,0.2882162,0.17860581818181817,-0.0855893917818592,-0.1096103654626501,-0.1096103818181818
+m,m1,arf,ARF,0.9912945326409913,1.0,0.9813719061719312,1.0,0.9813719117647058,1.0,0.008705467359008678,0.018628093828068848,0.01862808823529416
+m,m1,bayesnet,BayesNet,0.992602475033248,1.0,0.9800420009144419,1.0,0.9800419705882353,1.0,0.007397524966751945,0.019957999085558065,0.01995802941176472
+m,m1,ctgan,CTGAN,0.9810719260044722,1.0,0.9653938069704345,1.0,0.965393794117647,1.0,0.01892807399552776,0.03460619302956547,0.03460620588235297
+m,m1,realtabformer,RealTabFormer,0.8471066785975507,0.9722222222222222,0.7703004145955464,0.9583333333333334,0.7703004705882353,0.95833325,0.12511554362467148,0.18803291873778694,0.18803277941176477
+m,m1,tabbyflow,TabbyFlow,0.9656531372922158,0.9861111111111112,0.9455694745840733,0.9791666666666667,0.9455695000000001,0.979166625,0.02045797381889536,0.033597192082593486,0.033597124999999894
+m,m1,tabddpm,TabDDPM,0.8323346282499426,1.0,0.7261974789915966,1.0,0.7261974705882352,1.0,0.16766537175005736,0.2738025210084034,0.27380252941176475
+m,m1,tabdiff,TabDiff,0.995012929094234,1.0,0.9882839246809835,1.0,0.9882839411764704,1.0,0.004987070905766,0.011716075319016528,0.011716058823529552
+m,m1,tabpfgen,TabPFGen,0.9312809464983377,1.0,0.8950504951144337,1.0,0.8950504411764706,1.0,0.0687190535016623,0.10494950488556631,0.10494955882352941
+m,m1,tabsyn,TabSyn,0.9151286003173399,1.0,0.8673567924083053,1.0,0.8673568235294118,1.0,0.08487139968266011,0.13264320759169468,0.1326431764705882
+m,m1,tvae,TVAE,0.6771013672022533,0.8125,0.514290108756053,0.71875,0.5142900882352941,0.71875,0.13539863279774667,0.20445989124394703,0.2044599117647059
+m,m10,arf,ARF,0.7123540719907316,0.896476827092335,0.5362522364250092,0.778299972372433,0.5362522000000001,0.7783,0.18412275510160336,0.2420477359474238,0.24204779999999992
+m,m10,bayesnet,BayesNet,0.951505998488504,0.9486418283447489,0.9197526483267934,0.8836916132037126,0.9197526285714286,0.8836916153846155,-0.0028641701437551736,-0.03606103512308079,-0.036061013186813096
+m,m10,ctgan,CTGAN,0.709624181546891,0.888682523616734,0.5319793459552495,0.7549257759784076,0.5319793428571429,0.7549257692307693,0.17905834206984306,0.22294643002315806,0.22294642637362638
+m,m10,forestdiffusion,ForestDiffusion,0.6009614712919183,0.5904878916899143,0.24229968098331292,0.15709431652783742,0.24229968571428573,0.15709434615384615,-0.010473579602004013,-0.0852053644554755,-0.08520533956043957
+m,m10,realtabformer,RealTabFormer,0.9854324615097559,0.9932648799164636,0.9704627005779807,0.9817177166724679,0.9704626571428571,0.9817177307692309,0.007832418406707675,0.011255016094487202,0.01125507362637379
+m,m10,tabbyflow,TabbyFlow,0.8217784927843358,0.9287957146089083,0.7131164637936611,0.8732985424901186,0.7131164571428571,0.8732985384615385,0.10701722182457252,0.1601820786964575,0.1601820813186814
+m,m10,tabddpm,TabDDPM,0.4828794070511674,0.5505513862463075,0.008556112503480923,0.01536279205962464,0.008556114285714286,0.015362807692307693,0.06767197919514012,0.006806679556143717,0.006806693406593407
+m,m10,tabpfgen,TabPFGen,0.7134644522666399,0.8916197641345622,0.5402908468776733,0.7637362637362638,0.5402908571428572,0.7637362692307692,0.17815531186792233,0.22344541685859054,0.22344541208791202
+m,m10,tabsyn,TabSyn,0.8360310930288815,0.9287453108416517,0.73773185312303,0.8728782661852933,0.7377318285714285,0.8728783076923077,0.09271421781277023,0.13514641306226327,0.13514647912087918
+m,m10,tvae,TVAE,0.7072203606577264,0.8908867521367522,0.5293407342487098,0.7615384615384616,0.5293407142857143,0.7615384615384616,0.18366639147902575,0.23219772728975185,0.23219774725274733
+m,m11,arf,ARF,0.9172031385653676,0.8978021978021977,0.7613211951447246,0.6934065934065934,0.7613211904761905,0.693406576923077,-0.019400940763169894,-0.06791460173813124,-0.06791461355311357
+m,m11,bayesnet,BayesNet,0.9111763688831358,0.8981900452488687,0.743240886098029,0.6945701357466063,0.7432409047619049,0.694570153846154,-0.012986323634267105,-0.048670750351422654,-0.048670750915750904
+m,m11,ctgan,CTGAN,0.9180789716305808,0.8977068214804065,0.7551353874883286,0.6986538461538461,0.755135380952381,0.6986538461538461,-0.02037215015017435,-0.05648154133448258,-0.056481534798534905
+m,m11,forestdiffusion,ForestDiffusion,0.8361961148419481,0.7935222672064778,0.5588115588115589,0.46153846153846156,0.5588115714285715,0.46153846153846156,-0.04267384763547033,-0.0972730972730973,-0.09727310989010995
+m,m11,realtabformer,RealTabFormer,0.9818886921159647,0.973511094386978,0.9515036730945822,0.9251559189697447,0.9515036428571428,0.9251559230769231,-0.008377597728986719,-0.026347754124837475,-0.02634771978021977
+m,m11,tabbyflow,TabbyFlow,0.724848130079691,0.6580935265248926,0.3569330897804311,0.3077902551582256,0.35693309523809524,0.30779026923076924,-0.06675460355479834,-0.04914283462220548,-0.049142826007326
+m,m11,tabddpm,TabDDPM,0.7802882205513785,0.731934731934732,0.44581453634085216,0.3496503496503497,0.44581454761904765,0.34965034615384616,-0.048353488616646456,-0.09616418669050247,-0.0961642014652015
+m,m11,tabsyn,TabSyn,0.724848130079691,0.6615706833348629,0.3569330897804311,0.30726689980938143,0.35693309523809524,0.30726692307692305,-0.06327744674482805,-0.04966618997104966,-0.049666172161172195
+m,m11,tvae,TVAE,0.863734222354912,0.8243978243978243,0.6804946011842564,0.5827505827505828,0.6804946190476191,0.582750576923077,-0.039336397957087765,-0.09774401843367364,-0.09774404212454213
+m,m12,arf,ARF,0.9167237197684159,0.9187904174015284,0.8485673320711684,0.8781856261022928,0.848567380952381,0.878185625,0.002066697633112513,0.02961829403112437,0.029618244047618925
+m,m12,bayesnet,BayesNet,0.8828335414872089,0.9761316872427983,0.7656653147355608,0.9641975308641976,0.7656653333333333,0.964197625,0.09329814575558937,0.1985322161286368,0.19853229166666664
+m,m12,ctgan,CTGAN,0.5775100636417043,0.6900826838102824,0.3569113539659579,0.5351240257154235,0.35691138095238095,0.5351241250000001,0.11257262016857805,0.1782126717494656,0.1782127440476191
+m,m12,realtabformer,RealTabFormer,0.9436254693691764,0.937617632891468,0.8570308092619305,0.8482702764976958,0.8570308333333333,0.84827025,-0.006007836477708306,-0.008760532764234719,-0.008760583333333294
+m,m12,tabbyflow,TabbyFlow,0.8735657390043355,0.9574074074074074,0.7574845038002932,0.9361111111111111,0.7574845238095238,0.936111125,0.08384166840307183,0.17862660731081792,0.17862660119047624
+m,m12,tabsyn,TabSyn,0.8653459001479789,0.9582157554379777,0.7361557173476678,0.9373236331569665,0.7361557857142857,0.937323625,0.0928698552899988,0.20116791580929871,0.20116783928571425
+m,m12,tvae,TVAE,0.5588746878526029,0.6215011734747077,0.3203131655657523,0.43196222993367006,0.32031314285714285,0.43196225,0.06262648562210482,0.11164906436791777,0.11164910714285714
+m,m2,arf,ARF,0.9754578754578755,0.868995752071595,0.9571428571428571,0.8034663758681959,0.9571429,0.8034663846153846,-0.10646212338628058,-0.15367648127466116,-0.15367651538461546
+m,m2,bayesnet,BayesNet,0.9333333333333333,0.9202612891730294,0.9,0.8803418803418803,0.9,0.8803418846153845,-0.013072044160303942,-0.0196581196581197,-0.019658115384615482
+m,m2,ctgan,CTGAN,0.9333333333333333,0.8817997580659724,0.9,0.8226495726495726,0.9,0.8226495769230768,-0.05153357526736091,-0.07735042735042741,-0.07735042307692319
+m,m2,realtabformer,RealTabFormer,0.9872727272727273,0.9481116743616743,0.9709090909090909,0.9146253746253746,0.9709091000000001,0.9146253461538462,-0.039161052911052985,-0.05628371628371631,-0.05628375384615392
+m,m2,tabddpm,TabDDPM,0.5819865319865319,0.7831362393862394,0.22111111111111112,0.6136363636363636,0.22111109999999998,0.6136363461538461,0.2011497073997075,0.3925252525252525,0.3925252461538461
+m,m2,tabpfgen,TabPFGen,0.9334858329829014,0.9459038687707354,0.9000455996352029,0.9188051033856172,0.9000456,0.9188051153846154,0.012418035787833959,0.01875950375041424,0.0187595153846154
+m,m2,tabsyn,TabSyn,0.9122222222222222,0.8827469084562866,0.7588888888888888,0.8040627509708644,0.7588889,0.8040627307692307,-0.029475313765935596,0.0451738620819756,0.04517383076923076
+m,m2,tvae,TVAE,0.9333333333333333,0.8689792388830048,0.9,0.8034188034188035,0.9,0.8034188076923077,-0.06435409445032858,-0.09658119658119657,-0.09658119230769235
+m,m4,arf,ARF,0.8223923020639157,0.8099857566308658,0.6833333333333333,0.7142857142857143,0.6833333333333333,0.7142857142857143,-0.012406545433049865,0.030952380952380953,0.030952380952380953
+m,m4,bayesnet,BayesNet,0.7986542394234293,0.809781558090083,0.6851851851851851,0.7142857142857143,0.6851851777777778,0.7142857142857143,0.01112731866665373,0.029100529100529182,0.029100536507936536
+m,m4,ctgan,CTGAN,0.782716049382716,0.8412698412698413,0.674074074074074,0.7619047619047619,0.6740741777777778,0.7619048571428572,0.058553791887125284,0.08783068783068781,0.08783067936507938
+m,m4,forestdiffusion,ForestDiffusion,0.9506172839506173,1.0,0.9259259259259258,1.0,0.9259259111111111,1.0,0.04938271604938271,0.07407407407407418,0.07407408888888889
+m,m4,realtabformer,RealTabFormer,0.9938271604938271,1.0,0.9907407407407408,1.0,0.9907407333333335,1.0,0.006172839506172867,0.00925925925925919,0.009259266666666544
+m,m4,tabbyflow,TabbyFlow,0.9876543209876543,1.0,0.9814814814814814,1.0,0.9814814666666667,1.0,0.012345679012345734,0.0185185185185186,0.01851853333333331
+m,m4,tabddpm,TabDDPM,0.5188271604938272,0.6507936507936508,0.0,0.0,0.0,0.0,0.13196649029982366,0.0,0.0
+m,m4,tabdiff,TabDiff,1.0,1.0,1.0,1.0,1.0,1.0,0.0,0.0,0.0
+m,m4,tabpfgen,TabPFGen,0.7943913506473944,0.8097834317734363,0.6787878787878788,0.7142857142857143,0.6787878888888889,0.7142857142857143,0.015392081126041846,0.03549783549783547,0.035497825396825444
+m,m4,tabsyn,TabSyn,1.0,1.0,1.0,1.0,1.0,1.0,0.0,0.0,0.0
+m,m4,tvae,TVAE,0.7506172839506173,0.8412698412698413,0.625925925925926,0.7619047619047619,0.6259260222222223,0.7619048571428572,0.09065255731922395,0.13597883597883587,0.1359788349206349
+m,m5,arf,ARF,0.9036047802959567,0.9745762334831161,0.8126631864867159,0.9333440850647328,0.8126632,0.9333441538461538,0.07097145318715936,0.12068089857801689,0.12068095384615374
+m,m5,bayesnet,BayesNet,0.7057080610021786,0.9654745838956365,0.558562091503268,0.8964237516869095,0.5585621,0.8964237692307692,0.2597665228934578,0.3378616601836415,0.3378616692307692
+m,m5,ctgan,CTGAN,0.5080765639589169,0.7115112780985607,0.2583053221288515,0.5632302926361036,0.2583053,0.5632303461538462,0.20343471413964387,0.3049249705072521,0.30492504615384614
+m,m5,forestdiffusion,ForestDiffusion,0.7755820105820106,0.6490355476423588,0.6322619047619048,0.38553203332615094,0.6322618999999999,0.38553203846153844,-0.12654646293965177,-0.24672987143575387,-0.2467298615384615
+m,m5,realtabformer,RealTabFormer,0.9857142857142858,0.9755500909347063,0.9785714285714286,0.9461023485061946,0.9785714000000001,0.9461023846153847,-0.010164194779579416,-0.03246908006523408,-0.03246901538461544
+m,m5,tabbyflow,TabbyFlow,0.9230753968253967,0.9453671099568932,0.8725595238095238,0.8892152846293714,0.8725596,0.8892153076923077,0.02229171313149647,0.016655760819847543,0.01665570769230773
+m,m5,tabddpm,TabDDPM,0.5961377504218062,0.47518886246956427,0.11151315789473684,0.009971509971509971,0.1115132,0.009971500000000001,-0.1209488879522419,-0.10154164792322687,-0.10154170000000001
+m,m5,tabdiff,TabDiff,0.8673980703392468,0.9467089197352355,0.788888888888889,0.8800306053595528,0.7888889,0.8800306153846154,0.07931084939598876,0.09114171647066382,0.09114171538461535
+m,m5,tabpfgen,TabPFGen,0.7866666666666666,0.974958374958375,0.6799999999999999,0.9467157842157843,0.6799999999999999,0.9467158076923077,0.18829170829170838,0.26671578421578435,0.26671580769230774
+m,m5,tabsyn,TabSyn,0.9481559290382819,0.9609018976666035,0.9097338935574231,0.9153369506310682,0.9097339,0.9153369615384614,0.012745968628321558,0.005603057073645168,0.005603061538461462
+m,m5,tvae,TVAE,0.4840242763772175,0.6497488512343239,0.22222689075630253,0.47077277215256497,0.222227,0.4707727692307693,0.1657245748571064,0.24854588139626244,0.24854576923076926
+m,m6,arf,ARF,0.8951351250208631,0.7929382860795102,0.7211579465579824,0.5405882658264177,0.7211579142857143,0.5405882307692308,-0.1021968389413529,-0.18056968073156465,-0.1805696835164835
+m,m6,bayesnet,BayesNet,0.8703036646283696,0.7952488687782805,0.659425196245218,0.5486425339366516,0.6594251999999999,0.5486425384615384,-0.07505479585008912,-0.11078266230856637,-0.1107826615384615
+m,m6,ctgan,CTGAN,0.8050907585386267,0.7175574082644414,0.5752261259233161,0.4371206361710053,0.5752261428571429,0.4371206538461539,-0.0875333502741853,-0.13810548975231085,-0.138105489010989
+m,m6,forestdiffusion,ForestDiffusion,0.8003896503580321,0.7040821192058238,0.5781548648078466,0.41892113354217425,0.5781548857142856,0.4189211538461538,-0.09630753115220836,-0.15923373126567236,-0.15923373186813183
+m,m6,realtabformer,RealTabFormer,0.8735801324355718,0.7896885249826426,0.6691195711777973,0.5419808949220714,0.6691195714285714,0.5419808461538461,-0.08389160745292912,-0.12713867625572595,-0.12713872527472525
+m,m6,tabbyflow,TabbyFlow,0.8281618240846176,0.7474425141661803,0.5805849821726261,0.494451871657754,0.5805850285714286,0.4944518846153846,-0.08071930991843723,-0.0861331105148721,-0.08613314395604399
+m,m6,tabddpm,TabDDPM,0.67945950045923,0.6300599727070314,0.12040816326530612,0.14102564102564102,0.12040817142857141,0.14102565384615384,-0.04939952775219858,0.0206174777603349,0.020617482417582433
+m,m6,tabdiff,TabDiff,0.8358503197242555,0.7616420478861916,0.5975842958742601,0.5136163723488034,0.5975843428571428,0.5136163461538461,-0.07420827183806389,-0.08396792352545668,-0.08396799670329669
+m,m6,tabpfgen,TabPFGen,0.7432012631856956,0.6672288583542401,0.6144662808326269,0.4656361252068563,0.6144663714285714,0.46563611538461536,-0.07597240483145551,-0.14883015562577057,-0.14883025604395606
+m,m6,tabsyn,TabSyn,0.8488302155838462,0.756156826971597,0.610087541416045,0.5054752265278581,0.6100875714285715,0.5054752307692307,-0.09267338861224916,-0.1046123148881869,-0.10461234065934077
+m,m6,tvae,TVAE,0.6291843417213064,0.6389923139336886,0.35472559107305995,0.33118739606296166,0.3547256285714286,0.33118734615384615,0.009807972212382166,-0.023538195010098295,-0.023538282417582435
+m,m7,arf,ARF,0.9682539682539683,1.0,0.9047619047619048,1.0,0.4222222,1.0,0.031746031746031744,0.09523809523809523,0.5777778
+m,m7,bayesnet,BayesNet,0.9591836734693878,1.0,0.8775510204081632,1.0,0.4095238,1.0,0.04081632653061218,0.12244897959183676,0.5904762
+m,m7,ctgan,CTGAN,0.9444444444444444,0.917948717948718,0.869047619047619,0.8769230769230769,0.40555553333333333,0.8769230769230769,-0.0264957264957264,0.007875457875457959,0.4713675435897436
+m,m7,forestdiffusion,ForestDiffusion,0.7375661375661375,0.7119188503803888,0.2285714285714286,0.41403212172442944,0.10666666666666667,0.4140321153846154,-0.025647287185748713,0.18546069315300084,0.3073654487179487
+m,m7,realtabformer,RealTabFormer,1.0,1.0,1.0,1.0,0.4666666666666667,1.0,0.0,0.0,0.5333333333333333
+m,m7,tabbyflow,TabbyFlow,0.9591836734693878,1.0,0.8775510204081632,1.0,0.4095238,1.0,0.04081632653061218,0.12244897959183676,0.5904762
+m,m7,tabddpm,TabDDPM,0.7210884353741497,0.7948717948717949,0.16326530612244897,0.5384615384615384,0.07619046666666666,0.5384615384615384,0.07378335949764525,0.3751962323390895,0.46227107179487176
+m,m7,tabdiff,TabDiff,0.9682539682539683,1.0,0.9047619047619048,1.0,0.4222222,1.0,0.031746031746031744,0.09523809523809523,0.5777778
+m,m7,tabpfgen,TabPFGen,0.9333333333333333,0.9102564102564104,0.8714285714285713,0.8653846153846154,0.4066666666666666,0.8653846153846154,-0.023076923076922995,-0.0060439560439559115,0.4587179487179488
+m,m7,tabsyn,TabSyn,0.9591836734693878,1.0,0.8775510204081632,1.0,0.4095238,1.0,0.04081632653061218,0.12244897959183676,0.5904762
+m,m7,tvae,TVAE,0.8373015873015872,0.8346153846153845,0.7083333333333333,0.7519230769230769,0.3305555333333333,0.7519230769230769,-0.0026862026862026767,0.04358974358974366,0.4213675435897436
+m,m8,arf,ARF,0.9739429859343044,0.9890833881218497,0.9387655300026434,0.9672501643655489,0.9387655333333333,0.9672501923076923,0.01514040218754531,0.028484634362905492,0.028484658974358923
+m,m8,bayesnet,BayesNet,0.9493384642803988,0.9684418145956607,0.8914268626503807,0.9053254437869822,0.8914269333333333,0.9053254615384615,0.01910335031526189,0.013898581136601473,0.013898528205128158
+m,m8,ctgan,CTGAN,0.7483863806444452,0.8734850890423645,0.6199100182971151,0.7799674525680718,0.6199100000000001,0.7799674615384615,0.12509870839791937,0.16005743427095664,0.16005746153846145
+m,m8,forestdiffusion,ForestDiffusion,0.8821335120635043,0.9419246688983531,0.8126333211216775,0.8941661408766671,0.8126332666666667,0.8941661538461538,0.05979115683484881,0.0815328197549896,0.08153288717948715
+m,m8,realtabformer,RealTabFormer,0.9703673116210356,0.9914364388048599,0.9531193514088251,0.9748088169140801,0.9531194,0.9748088461538462,0.021069127183824254,0.021689465505254968,0.02168944615384627
+m,m8,tabddpm,TabDDPM,0.9407877425910272,0.9893644367328578,0.8556626738083765,0.9680933101985734,0.8556626666666667,0.9680933461538461,0.048576694141830634,0.1124306363901969,0.11243067948717944
+m,m8,tabdiff,TabDiff,0.9741398360590983,0.9916548438815644,0.933736597796271,0.9759766773936815,0.9337366,0.9759766923076924,0.017515007822466178,0.042240079597410496,0.04224009230769232
+m,m8,tabpfgen,TabPFGen,0.9824685534591194,0.9926926144031408,0.9737028301886793,0.9790899889584099,0.9737028,0.97909,0.010224060944021396,0.005387158769730616,0.005387200000000036
+m,m8,tabsyn,TabSyn,0.9799894755062563,0.9929225819914079,0.9460042520639399,0.9807920374721995,0.9460042000000001,0.9807920384615384,0.012933106485151669,0.03478778540825955,0.03478783846153832
+m,m8,tvae,TVAE,0.615487240756058,0.8470788420983602,0.4232308611340869,0.7381981044486092,0.423231,0.738198153846154,0.23159160134230217,0.3149672433145223,0.3149671538461539
+m,m9,arf,ARF,0.8192931756402942,0.9206031625156741,0.6874353124972876,0.8588241135418125,0.6874353125,0.8588241153846153,0.10130998687537984,0.17138880104452492,0.17138880288461533
+m,m9,bayesnet,BayesNet,0.8063052383267566,0.9321787349890471,0.6762058140142022,0.8647435897435898,0.67620578125,0.864743576923077,0.12587349666229053,0.18853777572938757,0.18853779567307705
+m,m9,ctgan,CTGAN,0.7972575728772457,0.9129477542470538,0.67051387810559,0.8358974358974359,0.670513875,0.8358974230769232,0.11569018136980802,0.16538355779184588,0.16538354807692313
+m,m9,forestdiffusion,ForestDiffusion,0.5036287598044986,0.48046921227396133,0.0,0.0,0.0,0.0,-0.023159547530537283,0.0,0.0
+m,m9,realtabformer,RealTabFormer,0.9257243917220848,0.981642486554694,0.8337453820753593,0.9453455198647507,0.83374540625,0.9453455384615383,0.05591809483260923,0.11160013778939137,0.1116001322115383
+m,m9,tabddpm,TabDDPM,0.8427936878331627,0.9409563020199135,0.6651956062030076,0.8791208791208791,0.665195625,0.8791208846153846,0.09816261418675087,0.21392527291787156,0.21392525961538467
+m,m9,tabpfgen,TabPFGen,0.7396760481842636,0.7729085102002199,0.587787271228168,0.6282117640813293,0.58778728125,0.6282117692307693,0.03323246201595631,0.0404244928531613,0.04042448798076925
+m,m9,tabsyn,TabSyn,0.8842110165308399,0.9507427359837253,0.7466331365229685,0.9084689187217277,0.746633125,0.9084689230769231,0.06653171945288539,0.16183578219875927,0.16183579807692305
+m,m9,tvae,TVAE,0.6180180685059712,0.7451839510266207,0.3957408149511625,0.5842517385995647,0.3957408125,0.5842517307692308,0.12716588252064953,0.18851092364840222,0.18851091826923078
+n,n1,arf,ARF,1.0,1.0,1.0,1.0,1.0,1.0,0.0,0.0,0.0
+n,n1,bayesnet,BayesNet,1.0,1.0,1.0,1.0,1.0,1.0,0.0,0.0,0.0
+n,n1,ctgan,CTGAN,1.0,1.0,1.0,1.0,1.0,1.0,0.0,0.0,0.0
+n,n1,forestdiffusion,ForestDiffusion,0.33370226895406746,0.36715307638100964,0.0,0.0,0.0,0.0,0.033450807426942175,0.0,0.0
+n,n1,realtabformer,RealTabFormer,1.0,1.0,1.0,1.0,1.0,1.0,0.0,0.0,0.0
+n,n1,tabbyflow,TabbyFlow,1.0,1.0,1.0,1.0,1.0,1.0,0.0,0.0,0.0
+n,n1,tabddpm,TabDDPM,0.6666666666666666,0.6666666666666666,0.0,0.0,0.0,0.0,0.0,0.0,0.0
+n,n1,tabpfgen,TabPFGen,1.0,1.0,1.0,1.0,1.0,1.0,0.0,0.0,0.0
+n,n1,tabsyn,TabSyn,1.0,1.0,1.0,1.0,1.0,1.0,0.0,0.0,0.0
+n,n1,tvae,TVAE,1.0,1.0,1.0,1.0,1.0,1.0,0.0,0.0,0.0
+n,n10,arf,ARF,0.9523809523809524,0.9020013802622499,0.9285714285714286,0.8530020703933748,0.9285715000000001,0.8530022173913043,-0.05037957211870259,-0.07556935817805377,-0.07556928260869578
+n,n10,bayesnet,BayesNet,0.9728937728937729,0.9387163561076605,0.9565934065934065,0.9080745341614908,0.9565933846153847,0.9080745217391304,-0.03417741678611241,-0.04851887243191577,-0.04851886287625429
+n,n10,ctgan,CTGAN,0.9523809523809524,0.9031055900621119,0.9285714285714286,0.8546583850931677,0.9285715000000001,0.8546585217391305,-0.049275362318840554,-0.07391304347826089,-0.07391297826086962
+n,n10,forestdiffusion,ForestDiffusion,0.3817990566331425,0.40680323013149033,0.0,0.0,0.0,0.0,0.025004173498347826,0.0,0.0
+n,n10,realtabformer,RealTabFormer,0.9751221001221002,0.9547964113181504,0.934981684981685,0.8752587991718426,0.93498175,0.8752589130434782,-0.020325688803949804,-0.05972288580984242,-0.0597228369565217
+n,n10,tabbyflow,TabbyFlow,0.9728327228327228,0.9572118702553485,0.9592490842490842,0.9358178053830228,0.9592490192307692,0.935817695652174,-0.015620852577374333,-0.023431278866061334,-0.02343132357859523
+n,n10,tabddpm,TabDDPM,0.9523809523809524,0.89648033126294,0.9285714285714286,0.8447204968944099,0.9285715000000001,0.844720652173913,-0.05590062111801242,-0.08385093167701874,-0.0838508478260871
+n,n10,tabpfgen,TabPFGen,0.9752136752136752,0.9594202898550724,0.9628205128205128,0.9391304347826087,0.9628205192307693,0.9391304347826087,-0.01579338535860275,-0.023690078037904128,-0.023690084448160564
+n,n10,tabsyn,TabSyn,0.6517094017094016,0.6195652173913043,0.0,0.0,0.0,0.0,-0.03214418431809729,0.0,0.0
+n,n10,tvae,TVAE,0.9792429792429793,0.9641131815044858,0.9688644688644689,0.9461697722567288,0.968864423076923,0.9461697391304348,-0.015129797738493544,-0.02269469660774004,-0.022694683946488214
+n,n11,arf,ARF,1.0,1.0,1.0,1.0,1.0,1.0,0.0,0.0,0.0
+n,n11,bayesnet,BayesNet,1.0,1.0,1.0,1.0,1.0,1.0,0.0,0.0,0.0
+n,n11,ctgan,CTGAN,1.0,1.0,1.0,1.0,1.0,1.0,0.0,0.0,0.0
+n,n11,forestdiffusion,ForestDiffusion,0.6666666666666666,0.36404626925221345,0.0,0.0,0.0,0.0,-0.3026203974144532,0.0,0.0
+n,n11,realtabformer,RealTabFormer,1.0,1.0,1.0,1.0,1.0,1.0,0.0,0.0,0.0
+n,n11,tabbyflow,TabbyFlow,1.0,1.0,1.0,1.0,1.0,1.0,0.0,0.0,0.0
+n,n11,tabddpm,TabDDPM,1.0,1.0,1.0,1.0,1.0,1.0,0.0,0.0,0.0
+n,n11,tabpfgen,TabPFGen,1.0,1.0,1.0,1.0,1.0,1.0,0.0,0.0,0.0
+n,n11,tabsyn,TabSyn,0.6666666666666666,0.6521739130434783,0.0,0.0,0.0,0.0,-0.01449275362318836,0.0,0.0
+n,n11,tvae,TVAE,1.0,1.0,1.0,1.0,1.0,1.0,0.0,0.0,0.0
+n,n12,arf,ARF,0.9010016144302093,0.984675363622732,0.7215829653068345,0.9540260908681961,0.5411871666666667,0.9540261304347826,0.08367374919252268,0.2324431255613616,0.4128389637681159
+n,n12,bayesnet,BayesNet,0.7402675131784512,0.8742248812104022,0.4064397024599341,0.6226746436312063,0.30482983333333336,0.622674652173913,0.13395736803195102,0.21623494117127218,0.3178448188405797
+n,n12,ctgan,CTGAN,0.7160708151220699,0.8799008773821543,0.3167787684227321,0.6666618632294566,0.237584,0.6666617826086957,0.16383006226008434,0.3498830948067245,0.42907778260869567
+n,n12,forestdiffusion,ForestDiffusion,0.3740798716076808,0.507360625355212,0.013071895424836602,0.0,0.009803916666666667,0.0,0.13328075374753118,-0.013071895424836602,-0.009803916666666667
+n,n12,tabbyflow,TabbyFlow,0.5577108054714975,0.52749286186444,0.007936507936507936,0.014492753623188404,0.0059524166666666675,0.014492739130434782,-0.03021794360705754,0.006556245686680468,0.008540322463768114
+n,n12,tabddpm,TabDDPM,0.7281930315969944,0.8394367095740094,0.34425516330776307,0.7139623026350715,0.25819133333333333,0.7139623478260869,0.11124367797701495,0.3697071393273085,0.4557710144927536
+n,n12,tabpfgen,TabPFGen,0.46672073118120216,0.840232118680822,0.09645345337847835,0.5743648343033354,0.07234016666666666,0.5743648260869565,0.3735113874996199,0.4779113809248571,0.5020246594202898
+n,n12,tabsyn,TabSyn,0.5721842112388065,0.5222184543671637,0.017094017094017096,0.0,0.0128205,0.0,-0.04996575687164273,-0.017094017094017096,-0.0128205
+n,n12,tvae,TVAE,0.7381143797166243,0.9099728808797983,0.37405368732778227,0.7518097239657189,0.28054025,0.7518098695652173,0.17185850116317403,0.3777560366379366,0.47126961956521735
+n,n14,arf,ARF,0.6080980130980131,0.6002956302799973,0.32904761904761903,0.16032322315870806,0.3290475,0.1603231923076923,-0.007802382818015818,-0.16872439588891097,-0.1687243076923077
+n,n14,bayesnet,BayesNet,0.44971139971139973,0.585854274170966,0.17456709956709957,0.11819653536197658,0.174567,0.11819642307692307,0.1361428744595663,-0.056370564205122986,-0.056370576923076926
+n,n14,ctgan,CTGAN,0.3888888888888889,0.5779943865760429,0.08333333333333334,0.09615384615384616,0.0833333,0.09615380769230769,0.18910549768715396,0.012820512820512817,0.012820507692307692
+n,n14,realtabformer,RealTabFormer,0.6713851848634457,0.6485137669955268,0.23147157190635453,0.1687521982275778,0.2314715,0.1687522307692308,-0.02287141786791891,-0.06271937367877672,-0.0627192692307692
+n,n14,tabddpm,TabDDPM,0.4591131091131091,0.6156065456138438,0.06666666666666667,0.07692307692307693,0.0666667,0.0769231153846154,0.15649343650073466,0.010256410256410262,0.0102564153846154
+n,n14,tabpfgen,TabPFGen,0.3888888888888889,0.577996023130807,0.08333333333333334,0.09615384615384616,0.0833333,0.09615380769230769,0.1891071342419181,0.012820512820512817,0.012820507692307692
+n,n14,tabsyn,TabSyn,0.5131146631146631,0.45462753176736975,0.23633866133866133,0.08819560519095354,0.2363385,0.08819553846153846,-0.058487131347293386,-0.14814305614770779,-0.14814296153846154
+n,n14,tvae,TVAE,0.3888888888888889,0.5737208823025386,0.08333333333333334,0.08974358974358974,0.0833333,0.08974357692307693,0.1848319934136497,0.006410256410256401,0.006410276923076927
+n,n15,arf,ARF,0.9535353535353536,0.964957264957265,0.9303030303030304,0.9474358974358974,0.9303030454545455,0.9474359230769231,0.011421911421911402,0.01713286713286699,0.017132877622377696
+n,n15,bayesnet,BayesNet,0.9312169312169313,0.9430864528425503,0.8968253968253969,0.9146296792638255,0.8968254545454546,0.9146297307692308,0.01186952162561905,0.017804282438428687,0.017804276223776228
+n,n15,ctgan,CTGAN,0.36534391534391536,0.3865282865282865,0.04801587301587302,0.07979242979242979,0.048015863636363634,0.07979246153846153,0.021184371184371165,0.03177655677655677,0.0317765979020979
+n,n15,realtabformer,RealTabFormer,0.9352813852813853,0.9388278388278388,0.9029220779220779,0.9082417582417582,0.9029220909090909,0.9082417692307693,0.003546453546453554,0.005319680319680331,0.005319678321678345
+n,n15,tabddpm,TabDDPM,0.4300745550745551,0.438466625966626,0.0,0.0,0.0,0.0,0.00839207089207089,0.0,0.0
+n,n15,tabpfgen,TabPFGen,0.9794612794612795,0.9763532763532764,0.9691919191919193,0.9645299145299145,0.9691919090909091,0.9645299230769232,-0.0031080031080030768,-0.004662004662004726,-0.00466198601398593
+n,n15,tabsyn,TabSyn,0.5080363805701548,0.4647594412507059,0.0,0.0,0.0,0.0,-0.04327693931944887,0.0,0.0
+n,n15,tvae,TVAE,0.8295815295815295,0.8279729760765838,0.7398268398268398,0.7373483490504766,0.7398269090909092,0.7373483846153845,-0.001608553504945709,-0.002478490776363196,-0.002478524475524657
+n,n16,arf,ARF,0.6666666666666666,0.33364341085271315,0.0,0.0,0.0,0.0,-0.3330232558139535,0.0,0.0
+n,n16,bayesnet,BayesNet,1.0,1.0,1.0,1.0,1.0,1.0,0.0,0.0,0.0
+n,n16,ctgan,CTGAN,1.0,1.0,1.0,1.0,1.0,1.0,0.0,0.0,0.0
+n,n16,tabddpm,TabDDPM,0.6666666666666666,0.6666666666666666,0.0,0.0,0.0,0.0,0.0,0.0,0.0
+n,n16,tabsyn,TabSyn,0.6666666666666666,0.6666666666666666,0.0,0.0,0.0,0.0,0.0,0.0,0.0
+n,n16,tvae,TVAE,0.6666666666666666,1.0,0.5,1.0,0.5,1.0,0.33333333333333337,0.5,0.5
+n,n18,arf,ARF,0.6700050109958255,0.7644132445343051,0.3566693329414802,0.6466198668014577,0.356669350877193,0.6466197499999999,0.0944082335384796,0.2899505338599775,0.28995039912280696
+n,n18,bayesnet,BayesNet,0.5443617454104309,0.8224998418211692,0.07711056390412634,0.6437200008269919,0.07711054385964912,0.64372,0.2781380964107383,0.5666094369228656,0.5666094561403509
+n,n18,ctgan,CTGAN,0.4327729185276111,0.5414054478039934,0.10847396326841328,0.18900102200309574,0.10847398245614036,0.18900100000000003,0.10863252927638228,0.08052705873468245,0.08052701754385967
+n,n18,tvae,TVAE,0.5137663441363588,0.4852984577521943,0.10300859233073384,0.13842816132899194,0.10300861403508771,0.138428125,-0.028467886384164476,0.0354195689982581,0.035419510964912304
+n,n19,arf,ARF,0.7144013807735785,0.7576288516021272,0.5303511296654921,0.3510676775879358,0.5303510857142857,0.3510676153846154,0.04322747082854872,-0.17928345207755636,-0.1792834703296703
+n,n19,bayesnet,BayesNet,0.7980552283531557,0.7630558457851979,0.6144322179145154,0.31814710039088434,0.6144322571428571,0.3181470769230769,-0.03499938256795776,-0.29628511752363107,-0.2962851802197802
+n,n19,ctgan,CTGAN,0.5722158786348696,0.7402198312642229,0.3472914253016624,0.3141574942008483,0.3472913714285714,0.3141574615384615,0.1680039526293533,-0.033133931100814085,-0.03313390989010989
+n,n19,tabddpm,TabDDPM,0.3529018219505704,0.41346920538352916,0.005024212316505549,0.04784810311315808,0.005024114285714286,0.047848192307692314,0.060567383432958755,0.04282389079665253,0.042824078021978025
+n,n19,tabpfgen,TabPFGen,0.8687033174711439,0.8219244343213208,0.7280837057299043,0.5129583783975995,0.7280837142857143,0.5129582692307693,-0.04677888314982315,-0.21512532733230483,-0.21512544505494502
+n,n19,tvae,TVAE,0.5789743801800057,0.6147904087744667,0.35731544773278884,0.16292607859357738,0.3573153142857143,0.16292615384615386,0.03581602859446098,-0.19438936913921145,-0.19438916043956042
+n,n3,arf,ARF,0.33898386452410223,0.529665331849216,0.002812469899063798,0.01954125417160755,0.0028125,0.019541304347826086,0.19068146732511382,0.01672878427254375,0.016728804347826087
+n,n3,bayesnet,BayesNet,0.33766164709885993,0.5231429949079749,0.0,0.0,0.0,0.0,0.18548134780911496,0.0,0.0
+n,n3,ctgan,CTGAN,0.3613222854225917,0.5202880584064418,0.03571428571428571,0.03933747412008281,0.03571425,0.03933747826086957,0.15896577298385006,0.0036231884057970967,0.003623228260869568
+n,n3,forestdiffusion,ForestDiffusion,0.40894186246776226,0.5782592202177687,0.10714285714285714,0.10455486542443064,0.10714275,0.10455478260869565,0.1693173577500064,-0.0025879917184264967,-0.0025879673913043466
+n,n3,realtabformer,RealTabFormer,0.9242534692752737,0.9230675848038941,0.8593566420513661,0.8085836633578084,0.859356625,0.8085836956521739,-0.0011858844713796834,-0.05077297869355768,-0.05077292934782607
+n,n3,tabbyflow,TabbyFlow,0.4218620461165135,0.6755368678248795,0.1198077940978774,0.41326734894054007,0.11980768750000001,0.41326739130434786,0.253674821708366,0.2934595548426627,0.2934597038043478
+n,n3,tabddpm,TabDDPM,0.44100952233487395,0.5907740904900404,0.0017712619827694048,0.021569299468062575,0.001771125,0.02156930434782609,0.14976456815516642,0.01979803748529317,0.01979817934782609
+n,n3,tabdiff,TabDiff,0.48821568628610873,0.7371291791363376,0.1462634321098565,0.4301547647270828,0.14626350000000002,0.43015486956521737,0.24891349285022885,0.2838913326172263,0.2838913695652173
+n,n3,tabpfgen,TabPFGen,0.4340109970679368,0.614301835426995,0.07511034963947824,0.08118130715664065,0.0751105,0.08118130434782608,0.18029083835905824,0.006070957517162412,0.0060708043478260865
+n,n3,tabsyn,TabSyn,0.37828448108957285,0.6463250471725958,0.047097223425356875,0.35589296535366105,0.0470970625,0.355893,0.2680405660830229,0.30879574192830417,0.3087959375
+n,n3,tvae,TVAE,0.3613222854225917,0.5009643869088573,0.03571428571428571,0.03209109730848861,0.03571425,0.03209108695652174,0.13964210148626555,-0.0036231884057971037,-0.0036231630434782613
+n,n4,arf,ARF,0.5541097194606933,0.5039827237821629,0.21905549229846252,0.2507146730171655,0.21905545614035085,0.250714625,-0.050126995678530406,0.03165918071870297,0.03165916885964917
+n,n4,bayesnet,BayesNet,0.49833214876527987,0.5021943573667711,0.14035087719298245,0.25,0.14035087719298245,0.25,0.003862208601491224,0.10964912280701755,0.10964912280701755
+n,n4,ctgan,CTGAN,0.44569770559347127,0.41886128624976887,0.0456140350877193,0.125,0.0456140350877193,0.125,-0.0268364193437024,0.0793859649122807,0.0793859649122807
+n,n4,realtabformer,RealTabFormer,0.8055002293857263,0.8022110640233104,0.5597200819267405,0.587362358736598,0.5597201754385965,0.587362375,-0.003289165362415858,0.027642276809857536,0.027642199561403458
+n,n4,tabbyflow,TabbyFlow,0.6292995849008425,0.5508275741083635,0.22598059225820044,0.2723040755246822,0.22598056140350875,0.272304,-0.07847201079247901,0.04632348326648178,0.04632343859649124
+n,n4,tabddpm,TabDDPM,0.44083259213768133,0.4388392857142857,0.048133153396311294,0.0,0.04813317543859649,0.0,-0.0019933064233956244,-0.048133153396311294,-0.04813317543859649
+n,n4,tabpfgen,TabPFGen,0.49823263513267124,0.5023315349197455,0.14035087719298245,0.25,0.14035087719298245,0.25,0.004098899787074228,0.10964912280701755,0.10964912280701755
+n,n4,tabsyn,TabSyn,0.5899592491839274,0.5138060722655794,0.22490764860996298,0.25408554877057044,0.22490763157894736,0.254085625,-0.076153176918348,0.02917790016060745,0.029177993421052623
+n,n4,tvae,TVAE,0.4366342373924148,0.41900050845747805,0.049778545935109,0.12531885546641597,0.049778526315789474,0.125318875,-0.017633728934936754,0.07554030953130697,0.07554034868421053
+n,n5,arf,ARF,0.7027436703775527,0.6700836991689372,0.44327988701393634,0.5032237306146599,0.44327991228070174,0.50322375,-0.032659971208615524,0.05994384360072352,0.05994383771929823
+n,n5,bayesnet,BayesNet,0.6177108511978816,0.6684904992205887,0.2656201731672557,0.500007263641118,0.26562019298245615,0.50000725,0.05077964802270707,0.2343870904738623,0.23438705701754387
+n,n5,ctgan,CTGAN,0.5996871305158156,0.6310032998580115,0.2335784874743018,0.44379677605066203,0.23357850877192982,0.443796875,0.03131616934219594,0.21021828857636024,0.21021836622807016
+n,n5,forestdiffusion,ForestDiffusion,0.46797444021157153,0.3356760079331851,0.007590859858842089,0.0,0.007590877192982457,0.0,-0.13229843227838645,-0.007590859858842089,-0.007590877192982457
+n,n5,realtabformer,RealTabFormer,0.7706436073803463,0.8225629873636423,0.363966738235617,0.5143573955572535,0.36396678947368416,0.5143575,0.05191937998329599,0.15039065732163648,0.15039071052631586
+n,n5,tabbyflow,TabbyFlow,0.6752698793844083,0.6659420028402353,0.32481295393648285,0.48429244475640276,0.32481292982456145,0.484292375,-0.009327876544172997,0.1594794908199199,0.15947944517543855
+n,n5,tabddpm,TabDDPM,0.37886848538454876,0.38748889352959953,0.013543142961477963,0.001893939393939394,0.013543157894736842,0.001894,0.008620408145050773,-0.011649203567538569,-0.011649157894736842
+n,n5,tabpfgen,TabPFGen,0.6403145015897745,0.6667644399648442,0.3262389213539785,0.49740932642487046,0.3262389122807018,0.497409375,0.026449938375069704,0.17117040507089198,0.1711704627192982
+n,n5,tabsyn,TabSyn,0.5761404931710918,0.3559017048077247,0.09264694201434613,0.0018463165491499978,0.09264692982456141,0.001846375,-0.22023878836336713,-0.09080062546519613,-0.09080055482456141
+n,n5,tvae,TVAE,0.5990774021209961,0.6190984351149365,0.23470212779423305,0.4259396331935192,0.2347021403508772,0.42593975,0.020021032993940313,0.19123750539928613,0.19123760964912281
+n,n6,arf,ARF,1.0,1.0,1.0,1.0,1.0,1.0,0.0,0.0,0.0
+n,n6,bayesnet,BayesNet,1.0,1.0,1.0,1.0,1.0,1.0,0.0,0.0,0.0
+n,n6,ctgan,CTGAN,1.0,1.0,1.0,1.0,1.0,1.0,0.0,0.0,0.0
+n,n6,forestdiffusion,ForestDiffusion,0.4602145345586766,0.3335951305713725,0.0,0.0,0.0,0.0,-0.1266194039873041,0.0,0.0
+n,n6,realtabformer,RealTabFormer,1.0,1.0,1.0,1.0,1.0,1.0,0.0,0.0,0.0
+n,n6,tabbyflow,TabbyFlow,1.0,1.0,1.0,1.0,1.0,1.0,0.0,0.0,0.0
+n,n6,tabddpm,TabDDPM,0.6578947368421053,0.6666666666666666,0.0,0.0,0.0,0.0,0.00877192982456132,0.0,0.0
+n,n6,tabdiff,TabDiff,1.0,1.0,1.0,1.0,1.0,1.0,0.0,0.0,0.0
+n,n6,tabpfgen,TabPFGen,1.0,1.0,1.0,1.0,1.0,1.0,0.0,0.0,0.0
+n,n6,tabsyn,TabSyn,0.6024227234753551,0.4666666666666666,0.0,0.0,0.0,0.0,-0.13575605680868846,0.0,0.0
+n,n6,tvae,TVAE,1.0,1.0,1.0,1.0,1.0,1.0,0.0,0.0,0.0
+n,n7,arf,ARF,0.9777777777777777,0.9366459627329193,0.9666666666666667,0.9049689440993788,0.9666666739130435,0.9049689565217391,-0.04113181504485841,-0.061697722567287894,-0.06169771739130436
+n,n7,bayesnet,BayesNet,0.9940763745111572,0.9797101449275363,0.9882677708764666,0.9695652173913044,0.9882677826086956,0.9695652173913044,-0.014366229583620882,-0.018702553485162143,-0.01870256521739122
+n,n7,ctgan,CTGAN,0.9717391304347827,0.9384057971014492,0.9576086956521739,0.907608695652174,0.9576086956521739,0.907608695652174,-0.03333333333333344,-0.04999999999999993,-0.04999999999999993
+n,n7,forestdiffusion,ForestDiffusion,0.5676322996601962,0.5275684786181144,0.0,0.0,0.0,0.0,-0.04006382104208184,0.0,0.0
+n,n7,realtabformer,RealTabFormer,0.9876409017713366,0.9809178743961352,0.981461352657005,0.9713768115942029,0.981461347826087,0.9713768260869565,-0.0067230273752013625,-0.0100845410628021,-0.010084521739130436
+n,n7,tabbyflow,TabbyFlow,0.9877415458937198,0.9722912353347136,0.9793478260869565,0.9527432712215321,0.9793478260869565,0.9527433043478261,-0.015450310559006142,-0.026604554865424346,-0.026604521739130416
+n,n7,tabddpm,TabDDPM,0.906561996779388,0.8752933057280884,0.8598429951690821,0.8129399585921324,0.859843,0.81294,-0.03126869105129959,-0.04690303657694972,-0.04690300000000003
+n,n7,tabpfgen,TabPFGen,0.9921497584541062,0.9942028985507246,0.9882246376811594,0.991304347826087,0.9882246304347827,0.991304347826087,0.002053140096618389,0.003079710144927583,0.003079717391304304
+n,n7,tabsyn,TabSyn,0.9863123993558776,0.9618357487922706,0.9794685990338164,0.9427536231884058,0.9794686086956522,0.9427536521739129,-0.024476650563606928,-0.03671497584541061,-0.036714956521739284
+n,n7,tvae,TVAE,0.9813520818955601,0.9673740510697032,0.9695048309178744,0.9460144927536231,0.9695048478260869,0.9460145217391305,-0.013978030825856913,-0.02349033816425128,-0.023490326086956448
+n,n8,arf,ARF,0.6251089324618736,0.5664541851692689,0.11659663865546217,0.01282051282051282,0.11659664285714287,0.0128205,-0.05865474729260467,-0.10377612583494936,-0.10377614285714287
+n,n8,bayesnet,BayesNet,0.5983196697482411,0.6850439287590125,0.019501133786848073,0.3557692307692308,0.01950111904761905,0.3557692307692308,0.08672425901077141,0.3362680969823827,0.33626811172161175
+n,n8,ctgan,CTGAN,0.6556831065759636,0.7074798261949099,0.15128968253968253,0.41025641025641024,0.15128954761904762,0.41025630769230764,0.051796719618946274,0.2589667277167277,0.25896676007326
+n,n8,realtabformer,RealTabFormer,0.6609279609279609,0.740481616280849,0.09523809523809523,0.45221407961177784,0.095238,0.45221403846153846,0.07955365535288805,0.3569759843736826,0.35697603846153847
+n,n8,tabbyflow,TabbyFlow,0.6825803825803826,0.7410641615674527,0.123015873015873,0.431588174615558,0.12301576190476189,0.43158815384615384,0.058483778987070134,0.30857230159968496,0.30857239194139197
+n,n8,tabsyn,TabSyn,0.6685931352598019,0.7886148886148886,0.14387464387464388,0.49734154734154734,0.1438745238095238,0.49734146153846154,0.12002175335508669,0.35346690346690346,0.3534669377289377
+n,n8,tvae,TVAE,0.576007326007326,0.6803316820479292,0.0,0.3659420289855072,0.0,0.36594203846153844,0.1043243560406032,0.3659420289855072,0.36594203846153844
+n,n9,arf,ARF,0.9023563689979781,0.9084463119555429,0.7322930579604596,0.7457716281743209,0.7322930714285715,0.7457716538461538,0.00608994295756482,0.01347857021386123,0.013478582417582374
+n,n9,bayesnet,BayesNet,0.9146991253680126,0.9062763056058766,0.7459832789800399,0.7195981475868606,0.7459832619047619,0.7195981153846154,-0.008422819762135991,-0.02638513139317933,-0.026385146520146563
+n,n9,ctgan,CTGAN,0.8618194849982179,0.8255844599197449,0.6203778351376499,0.5612496832573882,0.6203777857142857,0.5612496538461539,-0.03623502507847298,-0.05912815188026166,-0.05912813186813182
+n,n9,forestdiffusion,ForestDiffusion,0.5008791037833942,0.5348965594616358,0.0016052625670730338,0.0,0.001605261904761905,0.0,0.03401745567824166,-0.0016052625670730338,-0.001605261904761905
+n,n9,realtabformer,RealTabFormer,0.9503274172161063,0.9424421770539043,0.8524252530913204,0.8297140246206351,0.8524253333333334,0.8297140769230769,-0.007885240162201934,-0.022711228470685296,-0.02271125641025651
+n,n9,tabbyflow,TabbyFlow,0.8920611886177495,0.8932655281744917,0.7090654254677812,0.7010060298400634,0.7090654047619047,0.7010060384615385,0.001204339556742151,-0.008059395627717736,-0.008059366300366189
+n,n9,tabddpm,TabDDPM,0.9073333324979693,0.9092649083311402,0.7505623775414367,0.7486198819792915,0.7505624285714285,0.7486198846153846,0.0019315758331708288,-0.0019424955621452078,-0.001942543956043874
+n,n9,tabpfgen,TabPFGen,0.9276762493282755,0.9332255488617123,0.807785937989541,0.820109338892829,0.8077859761904762,0.8201093461538461,0.005549299533436747,0.012323400903287984,0.012323369963369912
+n,n9,tabsyn,TabSyn,0.6094770202643847,0.5376812840228072,0.026571024449383713,0.0,0.02657102380952381,0.0,-0.07179573624157742,-0.026571024449383713,-0.02657102380952381
+n,n9,tvae,TVAE,0.8615945954664943,0.8416538534778207,0.6395442934775143,0.6246397890459736,0.639544380952381,0.6246397307692307,-0.01994074198867357,-0.014904504431540677,-0.014904650183150236

evaluation/query_family/subgroup/shared_basis_stability_analysis/support3_prefix_advantage.csv

@@ -0,0 +1,4 @@
+dataset_prefix,paired_panel_count,support3_diff,key_diff,size_better_support3_share
+c,152,-0.015363644029073104,-0.010274899559989805,0.3223684210526316
+m,107,0.031793397956303,0.05606537668812528,0.6261682242990654
+n,139,0.02419863304939069,0.041224088576314835,0.4172661870503597

evaluation/query_family/subgroup/shared_basis_stability_analysis/support3_query_ecdf.tex

@@ -0,0 +1,49 @@
+\documentclass[tikz,border=4pt]{standalone}
+\usepackage{pgfplots}
+\pgfplotsset{compat=1.18}
+\definecolor{internalcolor}{HTML}{AA3377}
+\definecolor{sizecolor}{HTML}{009988}
+\begin{document}
+\begin{tikzpicture}
+\begin{axis}[
+width=14.8cm,
+height=8.7cm,
+xmin=0.0, xmax=1.0,
+ymin=0.0, ymax=1.02,
+xlabel={Support3 score},
+ylabel={Share of queries with score >= x},
+title={Support-side stability on a shared subgroup basis},
+ymajorgrids,
+grid style={draw=gray!20},
+major grid style={draw=gray!30},
+axis line style={draw=black!70},
+tick style={draw=black!70},
+legend style={draw=none, fill=none, font=\small, at={(0.98,0.98)}, anchor=north east},
+]
+\addplot[line width=1.9pt, color=internalcolor]
+table[x=score, y=internal_survival, col sep=comma]{support3_ecdf_curves.csv};
+\addlegendentry{Internal profile stability}
+\addplot[line width=1.9pt, color=sizecolor]
+table[x=score, y=size_survival, col sep=comma]{support3_ecdf_curves.csv};
+\addlegendentry{Subgroup size stability}
+\addplot[dashed, line width=1.0pt, color=internalcolor, opacity=0.85]
+coordinates {(0.777030,0) (0.777030,1.02)};
+\addplot[dashed, line width=1.0pt, color=sizecolor, opacity=0.85]
+coordinates {(0.818445,0) (0.818445,1.02)};
+\node[
+anchor=north west,
+draw=gray!55,
+fill=white,
+rounded corners=2pt,
+fill opacity=0.96,
+text opacity=1,
+align=left,
+font=\small
+] at (axis description cs:0.015,0.97) {
+Means: internal = 0.777, size = 0.818\\
+Medians: internal = 0.889, size = 1.000\\
+Low-tail share (<= 0.4): internal = 0.143, size = 0.129
+};
+\end{axis}
+\end{tikzpicture}
+\end{document}