Upload benchmark overall table assets

evaluation/tables/benchmark_overall_table/README.md

@@ -0,0 +1,37 @@
+# Benchmark Overall Table
+
+This folder contains the conference-style benchmark summary table for the evaluation section.
+
+## Goal
+
+Create a single-page LaTeX summary table that matches the paper-table aesthetic requested for the evaluation section while following the README's frozen paper-facing model roster.
+
+- model rows
+- grouped metric headers
+- top-3 color highlighting per metric column
+- compact `mean ± std` cells
+- direct PDF export for review
+- companion CSV exports for exact values
+
+## Files
+
+- `build_overall_benchmark_table.py`
+  - Rebuilds the table from the current evaluation assets.
+- `final/benchmark_overall_table_real.tex`
+  - Standalone LaTeX source for the current real-data table.
+- `final/benchmark_overall_table_real_model_summary.csv`
+  - Model-level summary values, standard deviations, counts, and top-3 ranks.
+- `final/benchmark_overall_table_real_dataset_level.csv`
+  - Dataset-level values used to compute the model summary table.
+- `final/benchmark_overall_table_real_sources.csv`
+  - Metric-to-source manifest for traceability.
+
+## Rebuild
+
+From the repo root:
+
+```powershell
+python Evaluation\benchmark_overall_table\build_overall_benchmark_table.py
+```
+
+Then compile the generated LaTeX file with the bundled `tectonic.exe` or another LaTeX engine.

evaluation/tables/benchmark_overall_table/build_overall_benchmark_table.py

@@ -0,0 +1,707 @@
+from __future__ import annotations
+
+import math
+import re
+from pathlib import Path
+
+import pandas as pd
+
+
+REPO_ROOT = Path(__file__).resolve().parents[2]
+EVAL_ROOT = REPO_ROOT / "Evaluation"
+OUT_DIR = EVAL_ROOT / "benchmark_overall_table" / "final"
+
+# README -> Figure Color Convention -> MODEL_COLORS
+PAPER_MODEL_ORDER = [
+    "real",
+    "arf",
+    "bayesnet",
+    "ctgan",
+    "forestdiffusion",
+    "realtabformer",
+    "tabbyflow",
+    "tabddpm",
+    "tabdiff",
+    "tabpfgen",
+    "tabsyn",
+    "tvae",
+]
+
+MODEL_LABELS = {
+    "real": "REAL",
+    "arf": "ARF",
+    "bayesnet": "BayesNet",
+    "ctgan": "CTGAN",
+    "forestdiffusion": "ForestDiffusion",
+    "realtabformer": "RealTabFormer",
+    "tabbyflow": "TabbyFlow",
+    "tabddpm": "TabDDPM",
+    "tabdiff": "TabDiff",
+    "tabpfgen": "TabPFGen",
+    "tabsyn": "TabSyn",
+    "tvae": "TVAE",
+}
+
+FINAL_BASENAME = "benchmark_overall_table_real"
+MODEL_ORDER_MAP = {model_id: idx for idx, model_id in enumerate(PAPER_MODEL_ORDER)}
+
+COMPUTED_METRICS = [
+    {
+        "key": "distance_overall",
+        "group": "Classical Fidelity",
+        "title": "Dist. overall $\\uparrow$",
+        "direction": "higher",
+        "reference_value": 1.0,
+        "highlight_column": True,
+        "source_kind": "distance_dataset_export",
+        "source_file": "Evaluation/distance/runs/*/datasets/*/distance_summary__*.csv",
+    },
+    {
+        "key": "jsd_distance",
+        "group": "Classical Fidelity",
+        "title": "JSD $\\downarrow$",
+        "direction": "lower",
+        "reference_value": 0.0,
+        "highlight_column": False,
+        "source_kind": "distance_dataset_export",
+        "source_file": "Evaluation/distance/runs/*/datasets/*/distance_summary__*.csv",
+    },
+    {
+        "key": "ks_distance",
+        "group": "Classical Fidelity",
+        "title": "KS $\\downarrow$",
+        "direction": "lower",
+        "reference_value": 0.0,
+        "highlight_column": False,
+        "source_kind": "distance_dataset_export",
+        "source_file": "Evaluation/distance/runs/*/datasets/*/distance_summary__*.csv",
+    },
+    {
+        "key": "tvd_distance",
+        "group": "Classical Fidelity",
+        "title": "TVD $\\downarrow$",
+        "direction": "lower",
+        "reference_value": 0.0,
+        "highlight_column": False,
+        "source_kind": "distance_dataset_export",
+        "source_file": "Evaluation/distance/runs/*/datasets/*/distance_summary__*.csv",
+    },
+    {
+        "key": "wasserstein_distance",
+        "group": "Classical Fidelity",
+        "title": "Wasserstein $\\downarrow$",
+        "direction": "lower",
+        "reference_value": 0.0,
+        "highlight_column": False,
+        "source_kind": "distance_dataset_export",
+        "source_file": "Evaluation/distance/runs/*/datasets/*/distance_summary__*.csv",
+    },
+    {
+        "key": "query_overall",
+        "group": "Query-centric Families",
+        "title": "Query overall $\\uparrow$",
+        "direction": "higher",
+        "reference_value": 1.0,
+        "highlight_column": True,
+        "source_kind": "derived_query_overall",
+        "source_file": "Derived from the five query-centric family dataset-level scores in build_overall_benchmark_table.py.",
+    },
+    {
+        "key": "subgroup_structure",
+        "group": "Query-centric Families",
+        "title": "Subgroup $\\uparrow$",
+        "direction": "higher",
+        "reference_value": 1.0,
+        "highlight_column": False,
+        "source_kind": "query_family_dataset_export",
+        "source_file": "Evaluation/query_fivepart_breakdown/subgroup_breakdown/data/dataset_model_scores.csv",
+    },
+    {
+        "key": "conditional_dependency_structure",
+        "group": "Query-centric Families",
+        "title": "Conditional $\\uparrow$",
+        "direction": "higher",
+        "reference_value": 1.0,
+        "highlight_column": False,
+        "source_kind": "query_family_dataset_export",
+        "source_file": "Evaluation/query_fivepart_breakdown/conditional_breakdown/data/dataset_model_scores.csv",
+    },
+    {
+        "key": "tail_breakdown",
+        "group": "Query-centric Families",
+        "title": "Tail $\\uparrow$",
+        "direction": "higher",
+        "reference_value": 1.0,
+        "highlight_column": False,
+        "source_kind": "query_family_dataset_export",
+        "source_file": "Evaluation/query_fivepart_breakdown/tail_breakdown/data/dataset_model_scores.csv",
+    },
+    {
+        "key": "missingness_structure",
+        "group": "Query-centric Families",
+        "title": "Missingness $\\uparrow$",
+        "direction": "higher",
+        "reference_value": 1.0,
+        "highlight_column": False,
+        "source_kind": "query_family_dataset_export",
+        "source_file": "Evaluation/query_fivepart_breakdown/missingness_breakdown/data/dataset_model_scores.csv",
+    },
+    {
+        "key": "cardinality_structure",
+        "group": "Query-centric Families",
+        "title": "Cardinality $\\uparrow$",
+        "direction": "higher",
+        "reference_value": 1.0,
+        "highlight_column": False,
+        "source_kind": "cardinality_dataset_export",
+        "source_file": "Evaluation/query_fivepart_breakdown/cardinality/data/cleaned_results.csv",
+    },
+]
+
+PLACEHOLDER_COLUMNS = [
+    {
+        "key": "train_time",
+        "group": "Cost",
+        "title": "Train time",
+        "placeholder": True,
+    },
+    {
+        "key": "generation_time",
+        "group": "Cost",
+        "title": "Gen. time",
+        "placeholder": True,
+    },
+]
+
+DISPLAY_COLUMNS = COMPUTED_METRICS + PLACEHOLDER_COLUMNS
+
+METRIC_SPECS = {metric["key"]: metric for metric in COMPUTED_METRICS}
+DISPLAY_SPECS = {column["key"]: column for column in DISPLAY_COLUMNS}
+FAMILY_KEYS = [
+    "subgroup_structure",
+    "conditional_dependency_structure",
+    "tail_breakdown",
+    "missingness_structure",
+    "cardinality_structure",
+]
+
+
+def ensure_out_dir() -> None:
+    OUT_DIR.mkdir(parents=True, exist_ok=True)
+
+
+def dataset_prefix(dataset_id: str) -> str:
+    return str(dataset_id)[0] if dataset_id else ""
+
+
+def dataset_sort_tuple(dataset_id: str) -> tuple[str, int, str]:
+    text = str(dataset_id)
+    match = re.match(r"([A-Za-z]+)(\d+)", text)
+    if match:
+        return match.group(1), int(match.group(2)), text
+    return text, 0, text
+
+
+def latex_escape(text: str) -> str:
+    replacements = {
+        "\\": r"\textbackslash{}",
+        "&": r"\&",
+        "%": r"\%",
+        "$": r"\$",
+        "#": r"\#",
+        "_": r"\_",
+        "{": r"\{",
+        "}": r"\}",
+        "~": r"\textasciitilde{}",
+        "^": r"\textasciicircum{}",
+    }
+    escaped = text
+    for src, dst in replacements.items():
+        escaped = escaped.replace(src, dst)
+    return escaped
+
+
+def fmt_num(value: float) -> str:
+    if abs(value) < 5e-4:
+        value = 0.0
+    return f"{value:.2f}"
+
+
+def normalize_long_metric(
+    df: pd.DataFrame,
+    metric_key: str,
+    value_col: str,
+    source_path: str,
+    source_kind: str,
+) -> pd.DataFrame:
+    subset = df.copy()
+    subset = subset[subset["model_id"].isin(PAPER_MODEL_ORDER)].copy()
+    subset["model_label"] = subset["model_id"].map(MODEL_LABELS)
+    subset["metric_key"] = metric_key
+    subset["metric_group"] = METRIC_SPECS[metric_key]["group"]
+    subset["metric_title"] = METRIC_SPECS[metric_key]["title"]
+    subset["metric_value"] = subset[value_col]
+    subset["source_kind"] = source_kind
+    subset["source_path"] = source_path
+    subset["row_kind"] = subset["model_id"].map(lambda model_id: "reference" if model_id == "real" else "synthetic")
+    return subset[
+        [
+            "dataset_id",
+            "dataset_prefix",
+            "model_id",
+            "model_label",
+            "metric_key",
+            "metric_group",
+            "metric_title",
+            "metric_value",
+            "source_kind",
+            "source_path",
+            "row_kind",
+        ]
+    ]
+
+
+def load_distance_long() -> pd.DataFrame:
+    frames = []
+    cols = [
+        "dataset_id",
+        "model_id",
+        "timestamp_utc",
+        "jensen_shannon_distance",
+        "kolmogorov_smirnov_distance",
+        "total_variation_distance",
+        "wasserstein_distance",
+        "overall_fidelity_score",
+    ]
+    for path in (EVAL_ROOT / "distance" / "runs").glob("*/*/distance_summary__*.csv"):
+        try:
+            df = pd.read_csv(path, usecols=cols, low_memory=False)
+        except Exception:
+            continue
+        frames.append(df)
+
+    if not frames:
+        raise FileNotFoundError("No distance dataset exports were found under Evaluation/distance/runs.")
+
+    df = pd.concat(frames, ignore_index=True)
+    df = df[df["model_id"].isin([model for model in PAPER_MODEL_ORDER if model != "real"])].copy()
+    df["timestamp_utc"] = pd.to_datetime(df["timestamp_utc"], errors="coerce")
+    df["dataset_prefix"] = df["dataset_id"].map(dataset_prefix)
+    df = df.sort_values(["dataset_id", "model_id", "timestamp_utc"]).drop_duplicates(
+        ["dataset_id", "model_id"], keep="last"
+    )
+
+    metric_map = [
+        ("distance_overall", "overall_fidelity_score"),
+        ("jsd_distance", "jensen_shannon_distance"),
+        ("ks_distance", "kolmogorov_smirnov_distance"),
+        ("tvd_distance", "total_variation_distance"),
+        ("wasserstein_distance", "wasserstein_distance"),
+    ]
+    long_frames = []
+    for metric_key, value_col in metric_map:
+        long_frames.append(
+            normalize_long_metric(
+                df[["dataset_id", "dataset_prefix", "model_id", value_col]].rename(columns={value_col: "metric_tmp"}),
+                metric_key=metric_key,
+                value_col="metric_tmp",
+                source_path="distance_dataset_export_dedup_latest",
+                source_kind="distance_dataset_export",
+            )
+        )
+    return pd.concat(long_frames, ignore_index=True)
+
+
+def load_family_long() -> pd.DataFrame:
+    frames = []
+
+    subgroup_path = EVAL_ROOT / "query_fivepart_breakdown" / "subgroup_breakdown" / "data" / "dataset_model_scores.csv"
+    subgroup_df = pd.read_csv(subgroup_path)
+    frames.append(
+        normalize_long_metric(
+            subgroup_df[["dataset_id", "dataset_prefix", "model_id", "subgroup_structure_score"]],
+            metric_key="subgroup_structure",
+            value_col="subgroup_structure_score",
+            source_path=str(subgroup_path.relative_to(REPO_ROOT)),
+            source_kind="query_family_dataset_export",
+        )
+    )
+
+    conditional_path = EVAL_ROOT / "query_fivepart_breakdown" / "conditional_breakdown" / "data" / "dataset_model_scores.csv"
+    conditional_df = pd.read_csv(conditional_path)
+    frames.append(
+        normalize_long_metric(
+            conditional_df[["dataset_id", "dataset_prefix", "model_id", "conditional_dependency_structure_score"]],
+            metric_key="conditional_dependency_structure",
+            value_col="conditional_dependency_structure_score",
+            source_path=str(conditional_path.relative_to(REPO_ROOT)),
+            source_kind="query_family_dataset_export",
+        )
+    )
+
+    tail_path = EVAL_ROOT / "query_fivepart_breakdown" / "tail_breakdown" / "data" / "dataset_model_scores.csv"
+    tail_df = pd.read_csv(tail_path)
+    frames.append(
+        normalize_long_metric(
+            tail_df[["dataset_id", "dataset_prefix", "model_id", "tail_breakdown_score"]],
+            metric_key="tail_breakdown",
+            value_col="tail_breakdown_score",
+            source_path=str(tail_path.relative_to(REPO_ROOT)),
+            source_kind="query_family_dataset_export",
+        )
+    )
+
+    missing_path = EVAL_ROOT / "query_fivepart_breakdown" / "missingness_breakdown" / "data" / "dataset_model_scores.csv"
+    missing_df = pd.read_csv(missing_path)
+    frames.append(
+        normalize_long_metric(
+            missing_df[["dataset_id", "dataset_prefix", "model_id", "missingness_structure_score"]],
+            metric_key="missingness_structure",
+            value_col="missingness_structure_score",
+            source_path=str(missing_path.relative_to(REPO_ROOT)),
+            source_kind="query_family_dataset_export",
+        )
+    )
+
+    cardinality_path = EVAL_ROOT / "query_fivepart_breakdown" / "cardinality" / "data" / "cleaned_results.csv"
+    cardinality_raw = pd.read_csv(
+        cardinality_path,
+        low_memory=False,
+        usecols=["dataset", "model", "official_cardinality_range_score"],
+    )
+    cardinality_df = (
+        cardinality_raw.groupby(["dataset", "model"], as_index=False)["official_cardinality_range_score"].mean()
+        .rename(
+            columns={
+                "dataset": "dataset_id",
+                "model": "model_id",
+                "official_cardinality_range_score": "cardinality_structure_score",
+            }
+        )
+    )
+    cardinality_df["dataset_prefix"] = cardinality_df["dataset_id"].map(dataset_prefix)
+    frames.append(
+        normalize_long_metric(
+            cardinality_df[["dataset_id", "dataset_prefix", "model_id", "cardinality_structure_score"]],
+            metric_key="cardinality_structure",
+            value_col="cardinality_structure_score",
+            source_path=str(cardinality_path.relative_to(REPO_ROOT)),
+            source_kind="cardinality_dataset_export",
+        )
+    )
+
+    family_long = pd.concat(frames, ignore_index=True)
+    return family_long[family_long["model_id"] != "real"].copy()
+
+
+def derive_query_overall_long(family_long: pd.DataFrame) -> pd.DataFrame:
+    query_long = (
+        family_long[family_long["metric_key"].isin(FAMILY_KEYS)]
+        .groupby(["dataset_id", "dataset_prefix", "model_id", "model_label"], as_index=False)
+        .agg(metric_value=("metric_value", "mean"))
+    )
+    query_long["metric_key"] = "query_overall"
+    query_long["metric_group"] = METRIC_SPECS["query_overall"]["group"]
+    query_long["metric_title"] = METRIC_SPECS["query_overall"]["title"]
+    query_long["source_kind"] = METRIC_SPECS["query_overall"]["source_kind"]
+    query_long["source_path"] = "derived_query_overall"
+    query_long["row_kind"] = "synthetic"
+    return query_long[
+        [
+            "dataset_id",
+            "dataset_prefix",
+            "model_id",
+            "model_label",
+            "metric_key",
+            "metric_group",
+            "metric_title",
+            "metric_value",
+            "source_kind",
+            "source_path",
+            "row_kind",
+        ]
+    ]
+
+
+def add_reference_rows(long_df: pd.DataFrame) -> pd.DataFrame:
+    frames = [long_df]
+    for metric in COMPUTED_METRICS:
+        metric_key = metric["key"]
+        metric_df = long_df[
+            (long_df["metric_key"] == metric_key) & long_df["metric_value"].notna()
+        ].copy()
+        if metric_df.empty:
+            continue
+
+        reference_rows = (
+            metric_df[["dataset_id", "dataset_prefix"]]
+            .drop_duplicates()
+            .assign(
+                model_id="real",
+                model_label="REAL",
+                metric_key=metric_key,
+                metric_group=metric["group"],
+                metric_title=metric["title"],
+                metric_value=metric["reference_value"],
+                source_kind="real_self_reference",
+                source_path="real_self_reference",
+                row_kind="reference",
+            )
+        )
+        frames.append(reference_rows)
+
+    combined = pd.concat(frames, ignore_index=True)
+    combined["model_order"] = combined["model_id"].map(MODEL_ORDER_MAP)
+    return combined.sort_values(
+        ["metric_key", "dataset_id", "model_order"],
+        key=lambda col: col.map(dataset_sort_tuple) if col.name == "dataset_id" else col,
+    ).reset_index(drop=True)
+
+
+def assemble_dataset_level_table() -> pd.DataFrame:
+    distance_long = load_distance_long()
+    family_long = load_family_long()
+    query_overall_long = derive_query_overall_long(family_long)
+    long_df = pd.concat([distance_long, family_long, query_overall_long], ignore_index=True)
+    long_df = long_df[long_df["metric_value"].notna()].copy()
+    long_df = add_reference_rows(long_df)
+    return long_df[
+        [
+            "dataset_id",
+            "dataset_prefix",
+            "model_id",
+            "model_label",
+            "metric_key",
+            "metric_group",
+            "metric_title",
+            "metric_value",
+            "source_kind",
+            "source_path",
+            "row_kind",
+            "model_order",
+        ]
+    ]
+
+
+def build_model_summary(dataset_level: pd.DataFrame) -> pd.DataFrame:
+    grouped = (
+        dataset_level.groupby(
+            ["model_id", "model_label", "row_kind", "metric_key", "metric_group", "metric_title"],
+            as_index=False,
+        )
+        .agg(
+            metric_mean=("metric_value", "mean"),
+            metric_std=("metric_value", "std"),
+            metric_count=("metric_value", "count"),
+        )
+    )
+    grouped.loc[grouped["metric_count"] <= 1, "metric_std"] = 0.0
+
+    wide = pd.DataFrame(
+        {
+            "model_id": PAPER_MODEL_ORDER,
+            "model_label": [MODEL_LABELS[model_id] for model_id in PAPER_MODEL_ORDER],
+            "row_kind": ["reference" if model_id == "real" else "synthetic" for model_id in PAPER_MODEL_ORDER],
+            "model_order": [MODEL_ORDER_MAP[model_id] for model_id in PAPER_MODEL_ORDER],
+        }
+    )
+
+    for metric in COMPUTED_METRICS:
+        metric_key = metric["key"]
+        metric_rows = grouped[grouped["metric_key"] == metric_key][
+            ["model_id", "metric_mean", "metric_std", "metric_count"]
+        ].rename(
+            columns={
+                "metric_mean": f"{metric_key}_mean",
+                "metric_std": f"{metric_key}_std",
+                "metric_count": f"{metric_key}_count",
+            }
+        )
+        wide = wide.merge(metric_rows, on="model_id", how="left")
+
+    for metric in COMPUTED_METRICS:
+        metric_key = metric["key"]
+        mean_col = f"{metric_key}_mean"
+        rank_col = f"{metric_key}_rank"
+        ascending = metric["direction"] == "lower"
+        wide[rank_col] = pd.NA
+        ranking = (
+            wide[(wide["row_kind"] == "synthetic") & wide[mean_col].notna()][["model_id", mean_col]]
+            .sort_values(mean_col, ascending=ascending)
+            .reset_index(drop=True)
+        )
+        for idx, row in ranking.head(3).iterrows():
+            wide.loc[wide["model_id"] == row["model_id"], rank_col] = idx + 1
+
+    for placeholder in PLACEHOLDER_COLUMNS:
+        key = placeholder["key"]
+        wide[f"{key}_mean"] = pd.NA
+        wide[f"{key}_std"] = pd.NA
+        wide[f"{key}_count"] = pd.NA
+        wide[f"{key}_rank"] = pd.NA
+
+    return wide.sort_values("model_order").reset_index(drop=True)
+
+
+def render_cell_text(column_key: str, mean_value: float | None, std_value: float | None, rank: int | None) -> str:
+    spec = DISPLAY_SPECS[column_key]
+    prefix = r"\cellcolor{OverallTint} " if spec.get("highlight_column") else ""
+
+    if spec.get("placeholder"):
+        return prefix + ""
+
+    if pd.isna(mean_value):
+        return prefix + r"\textit{N/A}"
+
+    std_numeric = 0.0 if pd.isna(std_value) else float(std_value)
+    body = f"{fmt_num(float(mean_value))}$_{{\\pm {fmt_num(std_numeric)}}}$"
+    rank_value = None if pd.isna(rank) else int(rank)
+    if rank_value == 1:
+        body = rf"{{\color{{FirstPlace}}\textbf{{{body}}}}}"
+    elif rank_value == 2:
+        body = rf"{{\color{{SecondPlace}}\textbf{{{body}}}}}"
+    elif rank_value == 3:
+        body = rf"{{\color{{ThirdPlace}}\textbf{{{body}}}}}"
+    return prefix + body
+
+
+def build_header_cells(group_name: str) -> list[dict[str, object]]:
+    return [column for column in DISPLAY_COLUMNS if column["group"] == group_name]
+
+
+def render_latex(summary: pd.DataFrame) -> str:
+    row_lines = []
+    for _, row in summary.iterrows():
+        cells = [latex_escape(str(row["model_label"]))]
+        for column in DISPLAY_COLUMNS:
+            key = column["key"]
+            cells.append(render_cell_text(key, row.get(f"{key}_mean"), row.get(f"{key}_std"), row.get(f"{key}_rank")))
+        row_lines.append(" & ".join(cells) + r" \\")
+
+    classical_cols = build_header_cells("Classical Fidelity")
+    query_cols = build_header_cells("Query-centric Families")
+    cost_cols = build_header_cells("Cost")
+    total_cols = 1 + len(DISPLAY_COLUMNS)
+    tabular_spec = "@{}l " + " ".join("c" for _ in DISPLAY_COLUMNS) + "@{}"
+
+    header_cells = []
+    for column in DISPLAY_COLUMNS:
+        title = column["title"]
+        if column.get("highlight_column"):
+            header_cells.append(rf"\cellcolor{{OverallTint}} {title}")
+        else:
+            header_cells.append(title)
+
+    return rf"""\documentclass[10pt]{{article}}
+\usepackage[a4paper,landscape,margin=0.60in]{{geometry}}
+\usepackage[T1]{{fontenc}}
+\usepackage[utf8]{{inputenc}}
+\usepackage{{newtxtext,newtxmath}}
+\usepackage{{booktabs}}
+\usepackage[table]{{xcolor}}
+\usepackage{{array}}
+\usepackage{{multirow}}
+\usepackage{{caption}}
+\usepackage{{microtype}}
+\usepackage{{graphicx}}
+\captionsetup{{font=small,labelfont=bf}}
+\definecolor{{FirstPlace}}{{HTML}}{{1397B8}}
+\definecolor{{SecondPlace}}{{HTML}}{{7B45E5}}
+\definecolor{{ThirdPlace}}{{HTML}}{{F28E2B}}
+\definecolor{{OverallTint}}{{HTML}}{{F8F1DA}}
+\definecolor{{RuleGray}}{{HTML}}{{C8CDD3}}
+\arrayrulecolor{{RuleGray}}
+\setlength{{\tabcolsep}}{{4.0pt}}
+\renewcommand{{\arraystretch}}{{1.12}}
+
+\begin{{document}}
+\thispagestyle{{empty}}
+
+\noindent{{\small\textit{{Conference-style benchmark summary for the evaluation section}}}}\\[-0.15em]
+\noindent\color{{RuleGray}}\rule{{\textwidth}}{{0.5pt}}
+
+\begin{{table}}[ht]
+\centering
+\caption{{Benchmark-wide summary of the frozen paper-facing model set specified in the README figure convention: 11 synthetic generators plus the \texttt{{REAL}} reference row. We report mean $\pm$ std across covered datasets using the current materialized evaluation exports. Lower is better for the four raw classical distance columns; higher is better for the two overall columns and the five query-centric family scores. The {{\color{{FirstPlace}} First}}, {{\color{{SecondPlace}} Second}}, and {{\color{{ThirdPlace}} Third}} best synthetic-model values in each column are highlighted with the same colors used in the table.}}
+\label{{tab:benchmark_overall_real}}
+\footnotesize
+\resizebox{{\textwidth}}{{!}}{{%
+\begin{{tabular}}{{{tabular_spec}}}
+\toprule
+\multirow{{2}}{{*}}{{\textbf{{Generator}}}} & \multicolumn{{{len(classical_cols)}}}{{c}}{{\textbf{{Classical Fidelity}}}} & \multicolumn{{{len(query_cols)}}}{{c}}{{\textbf{{Query-centric Families}}}} & \multicolumn{{{len(cost_cols)}}}{{c}}{{\textbf{{Cost}}}} \\
+\cmidrule(lr){{2-{1 + len(classical_cols)}}}
+\cmidrule(lr){{{2 + len(classical_cols)}-{1 + len(classical_cols) + len(query_cols)}}}
+\cmidrule(lr){{{2 + len(classical_cols) + len(query_cols)}-{total_cols}}}
+ & {" & ".join(header_cells)} \\
+\midrule
+{chr(10).join(row_lines)}
+\bottomrule
+\end{{tabular}}%
+}}
+
+\vspace{{0.45em}}
+\begin{{minipage}}{{0.95\linewidth}}
+\small
+\textit{{Note.}} The \texttt{{REAL}} row is a self-comparison reference row. For raw distance columns it is fixed to 0.00; for \texttt{{Dist. overall}}, \texttt{{Query overall}}, and the five family-score columns it is fixed to 1.00. The cost columns are intentionally left blank as placeholders for the training-time and generation-time statistics that will be added next.
+\end{{minipage}}
+\end{{table}}
+
+\end{{document}}
+"""
+
+
+def build_source_manifest() -> pd.DataFrame:
+    rows = []
+    for metric in COMPUTED_METRICS:
+        rows.append(
+            {
+                "metric_key": metric["key"],
+                "metric_group": metric["group"],
+                "metric_title": metric["title"],
+                "source_kind": metric["source_kind"],
+                "source_file": metric["source_file"],
+            }
+        )
+    for placeholder in PLACEHOLDER_COLUMNS:
+        rows.append(
+            {
+                "metric_key": placeholder["key"],
+                "metric_group": placeholder["group"],
+                "metric_title": placeholder["title"],
+                "source_kind": "placeholder",
+                "source_file": "Reserved placeholder column; real values will be added later.",
+            }
+        )
+    rows.append(
+        {
+            "metric_key": "real_reference_rows",
+            "metric_group": "Reference",
+            "metric_title": "REAL self-comparison rows",
+            "source_kind": "derived_reference",
+            "source_file": "Generated in build_overall_benchmark_table.py from the applicable dataset coverage of each metric.",
+        }
+    )
+    return pd.DataFrame(rows)
+
+
+def write_outputs(dataset_level: pd.DataFrame, summary: pd.DataFrame) -> None:
+    ensure_out_dir()
+    dataset_csv_path = OUT_DIR / f"{FINAL_BASENAME}_dataset_level.csv"
+    summary_csv_path = OUT_DIR / f"{FINAL_BASENAME}_model_summary.csv"
+    sources_csv_path = OUT_DIR / f"{FINAL_BASENAME}_sources.csv"
+    tex_path = OUT_DIR / f"{FINAL_BASENAME}.tex"
+
+    dataset_level.to_csv(dataset_csv_path, index=False)
+    summary.to_csv(summary_csv_path, index=False)
+    build_source_manifest().to_csv(sources_csv_path, index=False)
+    tex_path.write_text(render_latex(summary), encoding="utf-8")
+
+
+def main() -> None:
+    dataset_level = assemble_dataset_level_table()
+    summary = build_model_summary(dataset_level)
+    write_outputs(dataset_level, summary)
+
+
+if __name__ == "__main__":
+    main()

evaluation/tables/benchmark_overall_table/final/README.md

@@ -0,0 +1,18 @@
+# Final Assets
+
+- `benchmark_overall_table_real.tex`: standalone LaTeX source for the real-data table
+- `benchmark_overall_table_real.pdf`: compiled review PDF
+- `benchmark_overall_table_real.png`: rendered one-page preview image
+- `benchmark_overall_table_real_model_summary.csv`: model-level summary values, stds, counts, and ranks
+- `benchmark_overall_table_real_dataset_level.csv`: dataset-level metric values used for the table
+- `benchmark_overall_table_real_sources.csv`: source manifest for each displayed metric
+
+This version follows the README's frozen paper-facing model roster and uses the current materialized evaluation data.
+
+Current layout changes:
+
+- removed the mid-table gray banner row
+- added `Query overall` as the mean of the five query-centric family scores
+- switched the classical submetrics back to raw distances (`JSD`, `KS`, `TVD`, `Wasserstein`)
+- kept `Dist. overall` and `Query overall` visually emphasized with a light background
+- reserved two blank `Cost` columns for later training/generation time values

evaluation/tables/benchmark_overall_table/final/benchmark_overall_table_real.pdf

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:10cefa65ef8b0310f1d27d67e3ed83aab1e6705141650370b3e015a0b909bf9d
+size 27896

evaluation/tables/benchmark_overall_table/final/benchmark_overall_table_real.tex

@@ -0,0 +1,66 @@
+\documentclass[10pt]{article}
+\usepackage[a4paper,landscape,margin=0.60in]{geometry}
+\usepackage[T1]{fontenc}
+\usepackage[utf8]{inputenc}
+\usepackage{newtxtext,newtxmath}
+\usepackage{booktabs}
+\usepackage[table]{xcolor}
+\usepackage{array}
+\usepackage{multirow}
+\usepackage{caption}
+\usepackage{microtype}
+\usepackage{graphicx}
+\captionsetup{font=small,labelfont=bf}
+\definecolor{FirstPlace}{HTML}{1397B8}
+\definecolor{SecondPlace}{HTML}{7B45E5}
+\definecolor{ThirdPlace}{HTML}{F28E2B}
+\definecolor{OverallTint}{HTML}{F8F1DA}
+\definecolor{RuleGray}{HTML}{C8CDD3}
+\arrayrulecolor{RuleGray}
+\setlength{\tabcolsep}{4.0pt}
+\renewcommand{\arraystretch}{1.12}
+
+\begin{document}
+\thispagestyle{empty}
+
+\noindent{\small\textit{Conference-style benchmark summary for the evaluation section}}\\[-0.15em]
+\noindent\color{RuleGray}\rule{\textwidth}{0.5pt}
+
+\begin{table}[ht]
+\centering
+\caption{Benchmark-wide summary of the frozen paper-facing model set specified in the README figure convention: 11 synthetic generators plus the \texttt{REAL} reference row. We report mean $\pm$ std across covered datasets using the current materialized evaluation exports. Lower is better for the four raw classical distance columns; higher is better for the two overall columns and the five query-centric family scores. The {\color{FirstPlace} First}, {\color{SecondPlace} Second}, and {\color{ThirdPlace} Third} best synthetic-model values in each column are highlighted with the same colors used in the table.}
+\label{tab:benchmark_overall_real}
+\footnotesize
+\resizebox{\textwidth}{!}{%
+\begin{tabular}{@{}l c c c c c c c c c c c c c@{}}
+\toprule
+\multirow{2}{*}{\textbf{Generator}} & \multicolumn{5}{c}{\textbf{Classical Fidelity}} & \multicolumn{6}{c}{\textbf{Query-centric Families}} & \multicolumn{2}{c}{\textbf{Cost}} \\
+\cmidrule(lr){2-6}
+\cmidrule(lr){7-12}
+\cmidrule(lr){13-14}
+ & \cellcolor{OverallTint} Dist. overall $\uparrow$ & JSD $\downarrow$ & KS $\downarrow$ & TVD $\downarrow$ & Wasserstein $\downarrow$ & \cellcolor{OverallTint} Query overall $\uparrow$ & Subgroup $\uparrow$ & Conditional $\uparrow$ & Tail $\uparrow$ & Missingness $\uparrow$ & Cardinality $\uparrow$ & Train time & Gen. time \\
+\midrule
+REAL & \cellcolor{OverallTint} 1.00$_{\pm 0.00}$ & 0.00$_{\pm 0.00}$ & 0.00$_{\pm 0.00}$ & 0.00$_{\pm 0.00}$ & 0.00$_{\pm 0.00}$ & \cellcolor{OverallTint} 1.00$_{\pm 0.00}$ & 1.00$_{\pm 0.00}$ & 1.00$_{\pm 0.00}$ & 1.00$_{\pm 0.00}$ & 1.00$_{\pm 0.00}$ & 1.00$_{\pm 0.00}$ &  &  \\
+ARF & \cellcolor{OverallTint} {\color{SecondPlace}\textbf{0.91$_{\pm 0.13}$}} & {\color{ThirdPlace}\textbf{0.17$_{\pm 0.28}$}} & {\color{SecondPlace}\textbf{0.06$_{\pm 0.05}$}} & {\color{ThirdPlace}\textbf{0.15$_{\pm 0.27}$}} & {\color{SecondPlace}\textbf{0.02$_{\pm 0.02}$}} & \cellcolor{OverallTint} 0.49$_{\pm 0.19}$ & 0.43$_{\pm 0.17}$ & 0.48$_{\pm 0.21}$ & {\color{ThirdPlace}\textbf{0.41$_{\pm 0.28}$}} & 0.86$_{\pm 0.26}$ & 0.54$_{\pm 0.30}$ &  &  \\
+BayesNet & \cellcolor{OverallTint} 0.86$_{\pm 0.16}$ & 0.19$_{\pm 0.28}$ & 0.16$_{\pm 0.16}$ & 0.17$_{\pm 0.27}$ & 0.05$_{\pm 0.07}$ & \cellcolor{OverallTint} {\color{SecondPlace}\textbf{0.58$_{\pm 0.17}$}} & {\color{ThirdPlace}\textbf{0.44$_{\pm 0.17}$}} & {\color{ThirdPlace}\textbf{0.49$_{\pm 0.19}$}} & {\color{SecondPlace}\textbf{0.48$_{\pm 0.30}$}} & 0.82$_{\pm 0.25}$ & {\color{SecondPlace}\textbf{0.83$_{\pm 0.22}$}} &  &  \\
+CTGAN & \cellcolor{OverallTint} 0.81$_{\pm 0.13}$ & 0.22$_{\pm 0.22}$ & 0.27$_{\pm 0.18}$ & 0.20$_{\pm 0.22}$ & 0.08$_{\pm 0.07}$ & \cellcolor{OverallTint} {\color{ThirdPlace}\textbf{0.52$_{\pm 0.17}$}} & 0.41$_{\pm 0.18}$ & 0.46$_{\pm 0.20}$ & 0.35$_{\pm 0.22}$ & 0.81$_{\pm 0.27}$ & {\color{ThirdPlace}\textbf{0.77$_{\pm 0.22}$}} &  &  \\
+ForestDiffusion & \cellcolor{OverallTint} 0.59$_{\pm 0.25}$ & 0.73$_{\pm 0.38}$ & 0.13$_{\pm 0.12}$ & 0.73$_{\pm 0.38}$ & 0.03$_{\pm 0.04}$ & \cellcolor{OverallTint} 0.24$_{\pm 0.12}$ & 0.26$_{\pm 0.19}$ & 0.30$_{\pm 0.19}$ & 0.11$_{\pm 0.19}$ & {\color{SecondPlace}\textbf{0.96$_{\pm 0.00}$}} & 0.28$_{\pm 0.24}$ &  &  \\
+RealTabFormer & \cellcolor{OverallTint} {\color{FirstPlace}\textbf{0.91$_{\pm 0.15}$}} & {\color{FirstPlace}\textbf{0.15$_{\pm 0.24}$}} & {\color{FirstPlace}\textbf{0.05$_{\pm 0.09}$}} & {\color{FirstPlace}\textbf{0.14$_{\pm 0.24}$}} & {\color{FirstPlace}\textbf{0.01$_{\pm 0.01}$}} & \cellcolor{OverallTint} {\color{FirstPlace}\textbf{0.63$_{\pm 0.13}$}} & {\color{FirstPlace}\textbf{0.49$_{\pm 0.13}$}} & {\color{FirstPlace}\textbf{0.58$_{\pm 0.15}$}} & {\color{FirstPlace}\textbf{0.54$_{\pm 0.26}$}} & {\color{FirstPlace}\textbf{0.96$_{\pm 0.05}$}} & {\color{FirstPlace}\textbf{0.84$_{\pm 0.21}$}} &  &  \\
+TabbyFlow & \cellcolor{OverallTint} 0.78$_{\pm 0.26}$ & 0.34$_{\pm 0.38}$ & 0.09$_{\pm 0.06}$ & 0.32$_{\pm 0.38}$ & 0.04$_{\pm 0.04}$ & \cellcolor{OverallTint} 0.40$_{\pm 0.16}$ & 0.40$_{\pm 0.18}$ & 0.44$_{\pm 0.22}$ & 0.23$_{\pm 0.28}$ & 0.65$_{\pm 0.32}$ & 0.47$_{\pm 0.33}$ &  &  \\
+TabDDPM & \cellcolor{OverallTint} 0.56$_{\pm 0.29}$ & 0.59$_{\pm 0.37}$ & 0.35$_{\pm 0.32}$ & 0.57$_{\pm 0.38}$ & 0.23$_{\pm 0.20}$ & \cellcolor{OverallTint} 0.32$_{\pm 0.13}$ & 0.31$_{\pm 0.19}$ & 0.35$_{\pm 0.20}$ & 0.23$_{\pm 0.25}$ & 0.70$_{\pm 0.35}$ & 0.38$_{\pm 0.22}$ &  &  \\
+TabDiff & \cellcolor{OverallTint} {\color{ThirdPlace}\textbf{0.86$_{\pm 0.23}$}} & {\color{SecondPlace}\textbf{0.16$_{\pm 0.26}$}} & {\color{ThirdPlace}\textbf{0.07$_{\pm 0.06}$}} & {\color{SecondPlace}\textbf{0.14$_{\pm 0.25}$}} & 0.04$_{\pm 0.05}$ & \cellcolor{OverallTint} 0.39$_{\pm 0.14}$ & {\color{SecondPlace}\textbf{0.47$_{\pm 0.18}$}} & {\color{SecondPlace}\textbf{0.54$_{\pm 0.22}$}} & 0.11$_{\pm 0.21}$ & {\color{ThirdPlace}\textbf{0.89$_{\pm 0.11}$}} & 0.36$_{\pm 0.33}$ &  &  \\
+TabPFGen & \cellcolor{OverallTint} 0.86$_{\pm 0.14}$ & 0.23$_{\pm 0.26}$ & 0.10$_{\pm 0.09}$ & 0.21$_{\pm 0.26}$ & {\color{ThirdPlace}\textbf{0.02$_{\pm 0.02}$}} & \cellcolor{OverallTint} 0.49$_{\pm 0.18}$ & 0.42$_{\pm 0.19}$ & 0.49$_{\pm 0.21}$ & 0.36$_{\pm 0.27}$ & 0.77$_{\pm 0.30}$ & 0.67$_{\pm 0.28}$ &  &  \\
+TabSyn & \cellcolor{OverallTint} 0.68$_{\pm 0.26}$ & 0.60$_{\pm 0.43}$ & 0.07$_{\pm 0.08}$ & 0.60$_{\pm 0.43}$ & 0.02$_{\pm 0.03}$ & \cellcolor{OverallTint} 0.32$_{\pm 0.15}$ & 0.32$_{\pm 0.18}$ & 0.33$_{\pm 0.19}$ & 0.22$_{\pm 0.31}$ & 0.69$_{\pm 0.30}$ & 0.44$_{\pm 0.29}$ &  &  \\
+TVAE & \cellcolor{OverallTint} 0.80$_{\pm 0.18}$ & 0.26$_{\pm 0.28}$ & 0.25$_{\pm 0.17}$ & 0.24$_{\pm 0.28}$ & 0.08$_{\pm 0.08}$ & \cellcolor{OverallTint} 0.47$_{\pm 0.16}$ & 0.39$_{\pm 0.18}$ & 0.44$_{\pm 0.20}$ & 0.31$_{\pm 0.20}$ & 0.76$_{\pm 0.25}$ & 0.69$_{\pm 0.23}$ &  &  \\
+\bottomrule
+\end{tabular}%
+}
+
+\vspace{0.45em}
+\begin{minipage}{0.95\linewidth}
+\small
+\textit{Note.} The \texttt{REAL} row is a self-comparison reference row. For raw distance columns it is fixed to 0.00; for \texttt{Dist. overall}, \texttt{Query overall}, and the five family-score columns it is fixed to 1.00. The cost columns are intentionally left blank as placeholders for the training-time and generation-time statistics that will be added next.
+\end{minipage}
+\end{table}
+
+\end{document}

evaluation/tables/benchmark_overall_table/final/benchmark_overall_table_real_model_summary.csv

@@ -0,0 +1,13 @@
+model_id,model_label,row_kind,model_order,distance_overall_mean,distance_overall_std,distance_overall_count,jsd_distance_mean,jsd_distance_std,jsd_distance_count,ks_distance_mean,ks_distance_std,ks_distance_count,tvd_distance_mean,tvd_distance_std,tvd_distance_count,wasserstein_distance_mean,wasserstein_distance_std,wasserstein_distance_count,query_overall_mean,query_overall_std,query_overall_count,subgroup_structure_mean,subgroup_structure_std,subgroup_structure_count,conditional_dependency_structure_mean,conditional_dependency_structure_std,conditional_dependency_structure_count,tail_breakdown_mean,tail_breakdown_std,tail_breakdown_count,missingness_structure_mean,missingness_structure_std,missingness_structure_count,cardinality_structure_mean,cardinality_structure_std,cardinality_structure_count,distance_overall_rank,jsd_distance_rank,ks_distance_rank,tvd_distance_rank,wasserstein_distance_rank,query_overall_rank,subgroup_structure_rank,conditional_dependency_structure_rank,tail_breakdown_rank,missingness_structure_rank,cardinality_structure_rank,train_time_mean,train_time_std,train_time_count,train_time_rank,generation_time_mean,generation_time_std,generation_time_count,generation_time_rank
+real,REAL,reference,0,1.0,0.0,51,0.0,0.0,46,0.0,0.0,45,0.0,0.0,46,0.0,0.0,45,1.0,0.0,48.0,1.0,0.0,48,1.0,0.0,48.0,1.0,0.0,48.0,1.0,0.0,49.0,1.0,0.0,51.0,,,,,,,,,,,,,,,,,,,
+arf,ARF,synthetic,1,0.9055677339404236,0.1333881397455527,51,0.1686396570275395,0.2764149542556469,46,0.0645690738100766,0.049907547495445,45,0.1493555765718291,0.2729443793473305,46,0.0166554721368002,0.0184711747419083,45,0.643119,0.13158603867815455,48.0,0.684076,0.321546,48,0.636626,0.303188,48.0,0.414352,0.284643,48.0,0.859576,0.26205,49.0,0.620965,0.2964211631550952,51.0,2,3,2,3,2,,,,3,,,,,,,,,,
+bayesnet,BayesNet,synthetic,2,0.861035489268572,0.1593255314864186,51,0.1918693056218474,0.2800423569438904,46,0.1616965450041834,0.164071501507166,45,0.1722786777282608,0.2737268795663723,46,0.0531572132667621,0.0716268804514982,45,0.7219329999999999,0.1246361769348849,48.0,0.701987,0.311266,48,0.669044,0.295311,48.0,0.483312,0.304243,48.0,0.823078,0.25411,49.0,0.932244,0.2170797328116985,51.0,,,,,,2,2,2,2,,1,,,,,,,,
+ctgan,CTGAN,synthetic,3,0.8123072203677638,0.1340834517063039,47,0.219691845098548,0.2247743333595948,42,0.2725301758760839,0.1751933148359343,41,0.1984126249595633,0.2230617189002497,42,0.0769921254260524,0.0734695062634974,41,0.648922,0.12066862561173064,46.0,0.646701,0.331962,46,0.614117,0.290979,46.0,0.352728,0.222595,46.0,0.813005,0.269961,47.0,0.818059,0.2161663867131066,47.0,,,,,,,,,,,,,,,,,,,
+forestdiffusion,ForestDiffusion,synthetic,4,0.5925712871560507,0.2547762025891855,26,0.7322393215511053,0.3783942994030074,23,0.1299335791997103,0.1184333131601377,22,0.7279594738895535,0.3812325179548685,23,0.0263303206965842,0.0373561705738879,22,0.42582339999999996,0.11597133578791204,17.0,0.262739,0.366351,26,0.315026,0.328952,26.0,0.113008,0.187357,17.0,0.960127,0.0,17.0,0.478217,0.242298677579586,17.0,,,,,,,,,,2,,,,,,,,,
+realtabformer,RealTabFormer,synthetic,5,0.9119509283786016,0.1451802414589032,43,0.1497545658967733,0.239130788036234,39,0.0510948888254557,0.0871233952919869,37,0.135424083380648,0.2350291514824929,39,0.0099635786028176,0.0111886501183175,37,0.797364,0.0928887470574327,43.0,0.82744,0.229083,44,0.787366,0.227502,44.0,0.53603,0.25663,43.0,0.962799,0.051631,43.0,0.873185,0.2072367855702573,43.0,1,1,1,1,1,1,1,1,1,1,3,,,,,,,,
+tabbyflow,TabbyFlow,synthetic,6,0.7803487806830286,0.2647595121199843,31,0.3371916192967019,0.3759259151494235,28,0.092362699816346,0.0603765810471224,25,0.320967535858944,0.3775135314149407,28,0.0373122167085066,0.0417046379487634,25,0.5397299999999999,0.15291224702916306,31.0,0.598026,0.392722,31,0.553941,0.370259,31.0,0.230673,0.283262,31.0,0.653499,0.323997,31.0,0.662511,0.3286642196371467,31.0,,,,,,,,,,,,,,,,,,,
+tabddpm,TabDDPM,synthetic,7,0.5596512250922409,0.2897333382506756,38,0.5859943571319134,0.3743921778034383,33,0.3549974755634268,0.3205824124287394,32,0.5742949153608327,0.3819985280531603,33,0.2259653940422593,0.2011258303268372,32,0.42753680000000005,0.1439636066073658,37.0,0.346872,0.403939,37,0.367564,0.348157,37.0,0.225363,0.252863,37.0,0.699048,0.351026,38.0,0.498837,0.2158681052633974,38.0,,,,,,,,,,,,,,,,,,,
+tabdiff,TabDiff,synthetic,8,0.8642207911323687,0.2285573705357203,16,0.1575009106429631,0.2574055613404875,15,0.0657045997185684,0.0608648949933885,11,0.1398062775994466,0.2504622488314508,15,0.0357020079569338,0.0518114046735985,11,0.578308,0.13005529762782775,14.0,0.700413,0.364624,17,0.650655,0.355229,17.0,0.10507,0.212247,14.0,0.885162,0.106016,14.0,0.55024,0.3277697758253061,14.0,3,2,3,2,,,3,,,3,,,,,,,,,
+tabpfgen,TabPFGen,synthetic,9,0.8589910872863241,0.1395862266369712,38,0.2349422625839788,0.2638710801433853,34,0.0999793956797353,0.0877715947372478,33,0.212259362749018,0.2580033945635038,34,0.0171815687182266,0.0170723486371264,33,0.6730533999999999,0.13635447651457855,38.0,0.657221,0.361244,38,0.653108,0.309824,38.0,0.357444,0.265057,38.0,0.769764,0.296933,38.0,0.92773,0.2826676907414761,38.0,,,,,3,3,,3,,,2,,,,,,,,
+tabsyn,TabSyn,synthetic,10,0.680213628669073,0.2617339963545282,45,0.6021720821940022,0.4271114312428491,40,0.0673613378995536,0.0774427252296879,39,0.5950739028662382,0.4344746519219425,40,0.0212440715837567,0.0259327651665431,39,0.461801,0.15023032938986372,34.0,0.381235,0.404316,45,0.37237,0.359189,45.0,0.221835,0.309424,34.0,0.685331,0.300099,34.0,0.648234,0.2931518540972815,34.0,,,,,,,,,,,,,,,,,,,
+tvae,TVAE,synthetic,11,0.7975667886162016,0.175082620468212,49,0.2617806897807341,0.2789644437070004,44,0.2549778527752194,0.170972398606608,43,0.2366478995558346,0.2798111009116718,44,0.0808083449594111,0.0835453885451198,43,0.5881488,0.11704988793388368,47.0,0.59234,0.318457,47,0.574173,0.294348,47.0,0.311286,0.201439,47.0,0.755833,0.246376,47.0,0.707112,0.2306133053042871,48.0,,,,,,,,,,,,,,,,,,,

evaluation/tables/benchmark_overall_table/final/benchmark_overall_table_real_sources.csv

@@ -0,0 +1,15 @@
+metric_key,metric_group,metric_title,source_kind,source_file
+distance_overall,Classical Fidelity,Dist. overall $\uparrow$,distance_dataset_export,Evaluation/distance/runs/*/datasets/*/distance_summary__*.csv
+jsd_distance,Classical Fidelity,JSD $\downarrow$,distance_dataset_export,Evaluation/distance/runs/*/datasets/*/distance_summary__*.csv
+ks_distance,Classical Fidelity,KS $\downarrow$,distance_dataset_export,Evaluation/distance/runs/*/datasets/*/distance_summary__*.csv
+tvd_distance,Classical Fidelity,TVD $\downarrow$,distance_dataset_export,Evaluation/distance/runs/*/datasets/*/distance_summary__*.csv
+wasserstein_distance,Classical Fidelity,Wasserstein $\downarrow$,distance_dataset_export,Evaluation/distance/runs/*/datasets/*/distance_summary__*.csv
+query_overall,Query-centric Families,Query overall $\uparrow$,derived_query_overall_v2,Derived from the five v2 query-family model summaries in benchmark_overall_table_real_model_summary.csv.
+subgroup_structure,Query-centric Families,Subgroup $\uparrow$,query_family_model_summary_v2,Evaluation/query_fivepart_breakdown/subgroup_breakdown/final/v2/model_summary__v2.csv
+conditional_dependency_structure,Query-centric Families,Conditional $\uparrow$,query_family_model_summary_v2,Evaluation/query_fivepart_breakdown/conditional_breakdown/final/v2/model_summary__v2.csv
+tail_breakdown,Query-centric Families,Tail $\uparrow$,query_family_model_summary_v2,Evaluation/query_fivepart_breakdown/tail_breakdown/final/v2/model_summary__v2.csv
+missingness_structure,Query-centric Families,Missingness $\uparrow$,query_family_model_summary_v2,Evaluation/query_fivepart_breakdown/missingness_breakdown/final/v2/model_summary__v2.csv
+cardinality_structure,Query-centric Families,Cardinality $\uparrow$,query_family_model_summary_v2,Evaluation/query_fivepart_breakdown/cardinality/final/summary_by_model__v2.csv
+train_time,Cost,Train time,placeholder,Reserved placeholder column; real values will be added later.
+generation_time,Cost,Gen. time,placeholder,Reserved placeholder column; real values will be added later.
+real_reference_rows,Reference,REAL self-comparison rows,derived_reference,Generated in build_overall_benchmark_table.py from the applicable dataset coverage of each metric.