validate_dataset.py

#!/usr/bin/env python3
"""Validation & Diagnostic Visualization for iCCM CHW Triage Dataset."""

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

SCENARIOS = ['low_burden', 'moderate_burden', 'high_burden']


def load_scenarios(data_dir='data'):
    dfs = {}
    for sc in SCENARIOS:
        path = os.path.join(data_dir, f'iccm_{sc}.csv')
        if os.path.exists(path):
            dfs[sc] = pd.read_csv(path)
    return dfs


def make_report(dfs, output='validation_report.png'):
    fig, axes = plt.subplots(4, 2, figsize=(16, 22))
    fig.suptitle('iCCM Community Health Worker Triage — Validation Report',
                 fontsize=16, fontweight='bold', y=0.98)
    df = dfs.get('moderate_burden', list(dfs.values())[0])

    # Panel 1: True diagnosis distribution
    ax = axes[0, 0]
    diags = ['malaria', 'pneumonia', 'diarrhoea', 'mixed', 'other_febrile']
    colors_d = ['#e74c3c', '#3498db', '#f39c12', '#9b59b6', '#95a5a6']
    counts = [((df['true_diagnosis'] == d).sum()) for d in diags]
    ax.bar(range(5), counts, color=colors_d)
    ax.set_xticks(range(5))
    ax.set_xticklabels([d.replace('_', ' ').title() for d in diags], fontsize=8)
    for i, v in enumerate(counts):
        ax.text(i, v + 30, f'{v/len(df)*100:.1f}%', ha='center', fontsize=9)
    ax.set_ylabel('Count')
    ax.set_title('True Diagnosis (Moderate Burden)')

    # Panel 2: CHW action distribution
    ax = axes[0, 1]
    actions = df['chw_action'].value_counts()
    colors_a = {'treat_at_community': '#2ecc71', 'refer_urgently': '#e74c3c',
                'treat_and_refer': '#f39c12', 'refer': '#e67e22'}
    ax.bar(range(len(actions)), actions.values,
           color=[colors_a.get(a, '#95a5a6') for a in actions.index])
    ax.set_xticks(range(len(actions)))
    ax.set_xticklabels([a.replace('_', '\n') for a in actions.index], fontsize=7)
    for i, v in enumerate(actions.values):
        ax.text(i, v + 30, f'{v/len(df)*100:.1f}%', ha='center', fontsize=9)
    ax.set_ylabel('Count')
    ax.set_title('CHW Action')

    # Panel 3: Symptom prevalence
    ax = axes[1, 0]
    syms = ['fever', 'cough', 'fast_breathing', 'diarrhoea', 'any_danger_sign']
    vals = [df[s].mean() * 100 for s in syms]
    ax.barh(range(5), vals, color='#3498db', alpha=0.8)
    ax.set_yticks(range(5))
    ax.set_yticklabels([s.replace('_', ' ').title() for s in syms])
    for i, v in enumerate(vals):
        ax.text(v + 0.5, i, f'{v:.1f}%', va='center', fontsize=10)
    ax.set_xlabel('Prevalence (%)')
    ax.set_title('Symptom & Sign Prevalence')

    # Panel 4: Treatment given
    ax = axes[1, 1]
    treatments = ['act_given', 'amoxicillin_given', 'ors_given', 'zinc_given']
    t_vals = [df[t].mean() * 100 for t in treatments]
    colors_t = ['#e74c3c', '#3498db', '#f39c12', '#2ecc71']
    ax.bar(range(4), t_vals, color=colors_t, alpha=0.8)
    ax.set_xticks(range(4))
    ax.set_xticklabels(['ACT', 'Amoxicillin', 'ORS', 'Zinc'])
    for i, v in enumerate(t_vals):
        ax.text(i, v + 0.5, f'{v:.1f}%', ha='center', fontsize=10)
    ax.set_ylabel('% of all cases')
    ax.set_title('Treatment Given')

    # Panel 5: Cross-scenario diagnosis rates
    ax = axes[2, 0]
    x = np.arange(len(diags))
    width = 0.25
    for i, sc in enumerate(SCENARIOS):
        if sc not in dfs:
            continue
        d = dfs[sc]
        rates = [(d['true_diagnosis'] == diag).mean() * 100 for diag in diags]
        ax.bar(x + i * width, rates, width, label=sc.replace('_', ' ').title(), alpha=0.8)
    ax.set_xticks(x + width)
    ax.set_xticklabels([d.replace('_', ' ').title() for d in diags], fontsize=7)
    ax.set_ylabel('Prevalence (%)')
    ax.set_title('Diagnosis Rates Across Scenarios')
    ax.legend(fontsize=8)

    # Panel 6: MUAC distribution
    ax = axes[2, 1]
    ax.hist(df['muac_cm'], bins=50, color='#9b59b6', alpha=0.7, edgecolor='white')
    ax.axvline(11.5, color='red', ls='--', lw=1.5, label='SAM <11.5')
    ax.axvline(12.5, color='orange', ls='--', lw=1.5, label='MAM <12.5')
    ax.set_xlabel('MUAC (cm)')
    ax.set_title(f'MUAC Distribution (SAM={( df["nutrition_status"]=="SAM").mean()*100:.1f}%)')
    ax.legend(fontsize=8)

    # Panel 7: Age distribution by diagnosis
    ax = axes[3, 0]
    for diag, color in [('malaria', '#e74c3c'), ('pneumonia', '#3498db'), ('diarrhoea', '#f39c12')]:
        sub = df[df['true_diagnosis'] == diag]['age_months']
        ax.hist(sub, bins=30, alpha=0.5, color=color, label=diag.title(), edgecolor='white')
    ax.set_xlabel('Age (months)')
    ax.set_title('Age Distribution by Diagnosis')
    ax.legend(fontsize=9)

    # Panel 8: Referral rate across scenarios
    ax = axes[3, 1]
    ref_rates = []
    ds_rates = []
    sam_rates = []
    for sc in SCENARIOS:
        if sc not in dfs:
            continue
        d = dfs[sc]
        ref_rates.append(np.mean(['refer' in str(a) for a in d['chw_action']]) * 100)
        ds_rates.append(d['any_danger_sign'].mean() * 100)
        sam_rates.append((d['nutrition_status'] == 'SAM').mean() * 100)
    x = np.arange(len(SCENARIOS))
    ax.bar(x - 0.2, ref_rates, 0.2, label='Referral Rate', color='#e74c3c', alpha=0.8)
    ax.bar(x, ds_rates, 0.2, label='Danger Signs', color='#f39c12', alpha=0.8)
    ax.bar(x + 0.2, sam_rates, 0.2, label='SAM', color='#9b59b6', alpha=0.8)
    ax.set_xticks(x)
    ax.set_xticklabels([s.replace('_', '\n').title() for s in SCENARIOS], fontsize=8)
    ax.set_ylabel('%')
    ax.set_title('Referral, Danger Signs, SAM Across Scenarios')
    ax.legend(fontsize=8)

    plt.tight_layout(rect=[0, 0, 1, 0.97])
    plt.savefig(output, dpi=150, bbox_inches='tight')
    print(f'Saved validation report to {output}')
    plt.close()


if __name__ == '__main__':
    dfs = load_scenarios()
    if dfs:
        make_report(dfs)