grouped_multirow_latex

Render a grouped LaTeX table with two-level row headers (e.g., Domain and Task) and flat method columns (e.g., MLP, Modality, Component).

• Input must be a flat DataFrame with (row1, row2, column, value) format.

• Each value should be a float or a list of floats, from which mean ± std or stderr is computed.

• Bolds the best (min or max) per row.

---

🧾 Required LaTeX packages / commands

• \usepackage{booktabs}

• \usepackage{multirow}

• \newcommand{\highlightcolor}[1]{\colorbox[HTML]{bae6fb}{\textbf{#1}}}

---

📥 Arguments

Name	Type	Required	Description
df	pd.DataFrame	✅	DataFrame with columns for row1, row2, column, and values (as lists or floats).
row1	str	✅	Column name for the outer row grouping (e.g., Domain).
row2	str	✅	Column name for the row label (e.g., Task).
col	str	✅	Column name representing the method axis (e.g., Method).
value_column	str	✅	Column name containing scalar or list-of-floats to summarize.
highlight	str	❌	‘min’ or ‘max’ to bold best value per row.
stderr	bool	❌	Use standard error instead of std when formatting.
caption	str	❌	LaTeX caption to display below the table.
label	str	❌	Optional LaTeX label for referencing.

---

📦 Example Output

Click to show example code

import numpy as np
import pandas as pd
from swizz import table

np.random.seed(0)

domains = ["HalfCheetah", "Hopper", "Walker2d", "Ant"]
tasks = ["IL (↑) [1]", "Off-RL (↑) [1]", "Sensor failure (↑) [11]", "Dynamics change (↑) [4]"]
methods = ["MLP", "Modality", "Component"]

rows = []
for domain in domains:
    for task in tasks:
        for method in methods:
            values = np.round(np.random.normal(loc=1.0, scale=0.05, size=5), 3).tolist()
            rows.append({
                "Domain": domain,
                "Task": task,
                "Method": method,
                "score": values
            })

df = pd.DataFrame(rows)

latex = table(
    "grouped_multirow_latex",
    df=df,
    row1="Domain",
    row2="Task",
    col="Method",
    value_column="score",
    highlight="max",
    stderr=False,
    caption="Expert-normalized returns across domains and methods.",
    label="tab:tokenization_comparison"
)

print(latex)