per_feature_agreement.py

"""Generate the per-feature agreement figure used in the manuscript.

This is a cleaned-up version of the relevant cells from irr_kaileigh.ipynb.
Running this script saves `per_feature_agreement_kappa_vs_percent.png`
next to the script.
"""

from __future__ import annotations

from pathlib import Path
from typing import Optional, Union

import matplotlib.pyplot as plt
import numpy as np
import pandas as pd

BASE_DIR = Path(__file__).resolve().parent
DATA_DIR = BASE_DIR / 'data'
OUTPUT_PATH = BASE_DIR / 'per_feature_agreement_kappa_vs_percent.png'


def load_feature_summary(csv_path: Optional[Union[str, Path]] = None) -> pd.DataFrame:
    """Load the pre-computed per-feature IRR summary table."""
    if csv_path is None:
        csv_path = DATA_DIR / 'feature_irr_results.csv'
    df = pd.read_csv(csv_path)
    df['FeatureID'] = df['FeatureID'].str.strip()
    return df


def _configure_matplotlib() -> None:
    plt.rcParams.update({
        'font.family': 'DejaVu Sans',
        'axes.edgecolor': 'black',
        'axes.facecolor': 'white',
        'xtick.color': 'black',
        'ytick.color': 'black',
        'axes.titleweight': 'bold'
    })


def plot_per_feature_agreement(df: pd.DataFrame, output_path: Path = OUTPUT_PATH) -> None:
    """Reproduce the κ vs. % agreement plot."""
    _configure_matplotlib()

    fig, ax_k = plt.subplots(figsize=(7, 5))
    ypos = np.arange(len(df))
    ax_k.set_yticks(ypos)
    ax_k.set_yticklabels(df['FeatureID'])

    fontsize_label = 10
    fontsize_title = 12

    ax_k.errorbar(
        df['Mean Kappa'],
        ypos - 0.12,
        xerr=df['SD Kappa'],
        fmt='o',
        color='black',
        markersize=7,
        elinewidth=1.8,
        capsize=3,
        label='κ (mean ± SD)'
    )

    ax_k.errorbar(
        df['Mean %Agree'],
        ypos + 0.12,
        xerr=df['SD %Agree'],
        fmt='s',
        mfc='none',
        mec='#C00000',
        color='#C00000',
        markersize=7,
        elinewidth=1.8,
        capsize=3,
        label='Agreement (mean ± SD)'
    )

    ax_k.set_xlim(0, 1)
    ax_k.set_xlabel('Mean Value (κ and Agreement)', fontsize=fontsize_label)
    ax_k.set_ylabel('Feature ID', fontsize=fontsize_label)
    ax_k.set_ylim(-0.6, len(ypos) - 0.4)

    for spine in ax_k.spines.values():
        spine.set_linewidth(1.2)
    ax_k.tick_params(axis='both', labelsize=fontsize_label)

    ax_k.grid(axis='y', color='#b5b5b5', linewidth=0.7)
    ax_k.grid(axis='x', color='#dcdcdc', linewidth=0.7, linestyle='--', alpha=0.6)

    ax_k.set_title('Per-Feature Agreement', fontsize=fontsize_title, pad=18)

    handles, labels = ax_k.get_legend_handles_labels()
    fig.legend(handles, labels, loc='upper left', bbox_to_anchor=(0.02, 0.98),
               frameon=False, fontsize=fontsize_label)

    plt.tight_layout()
    fig.savefig(output_path, dpi=300, bbox_inches='tight')
    plt.close(fig)
    print(f"Saved figure -> {output_path}")


def main() -> None:
    df_summary = load_feature_summary()
    plot_per_feature_agreement(df_summary)


if __name__ == '__main__':
    main()