validate_dataset.py

#!/usr/bin/env python3
"""Validation & Diagnostic Visualization for Adolescent SRH Dataset."""

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

SCENARIOS = ['youth_friendly_clinic', 'public_health_facility', 'rural_limited_access']


def load_scenarios(data_dir='data'):
    dfs = {}
    for sc in SCENARIOS:
        path = os.path.join(data_dir, f'adolescent_srh_{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('Adolescent Sexual & Reproductive Health — Validation Report',
                 fontsize=16, fontweight='bold', y=0.98)
    df = dfs.get('public_health_facility', list(dfs.values())[0])

    # Panel 1: Contraceptive use across scenarios
    ax = axes[0, 0]
    x = np.arange(len(SCENARIOS))
    sa_dfs = {sc: dfs[sc][dfs[sc]['sexually_active'] == 1] for sc in SCENARIOS if sc in dfs}
    mcpr = [sa_dfs[sc]['using_modern_contraceptive'].mean() * 100 for sc in SCENARIOS if sc in sa_dfs]
    condom = [sa_dfs[sc]['condom_last_sex'].mean() * 100 for sc in SCENARIOS if sc in sa_dfs]
    width = 0.3
    ax.bar(x - width/2, mcpr, width, label='mCPR', color='#2ecc71', alpha=0.8)
    ax.bar(x + width/2, condom, width, label='Condom Use', color='#3498db', alpha=0.8)
    ax.set_xticks(x)
    ax.set_xticklabels(['Youth Clinic', 'Public Facility', 'Rural'], fontsize=8)
    ax.set_ylabel('Percentage (%)')
    ax.set_title('Contraceptive Use (Sexually Active)')
    ax.legend(fontsize=8)

    # Panel 2: Pregnancy rates across scenarios (females)
    ax = axes[0, 1]
    preg_curr = []
    preg_ever = []
    for sc in SCENARIOS:
        if sc in dfs:
            f = dfs[sc][dfs[sc]['sex'] == 'F']
            preg_curr.append(f['currently_pregnant'].mean() * 100)
            preg_ever.append(f['ever_pregnant'].mean() * 100)
    ax.bar(x - width/2, preg_curr, width, label='Currently Pregnant', color='#e74c3c', alpha=0.8)
    ax.bar(x + width/2, preg_ever, width, label='Ever Pregnant', color='#f39c12', alpha=0.8)
    ax.set_xticks(x)
    ax.set_xticklabels(['Youth Clinic', 'Public Facility', 'Rural'], fontsize=8)
    ax.set_ylabel('Percentage (%)')
    ax.set_title('Pregnancy Rates (Females)')
    ax.legend(fontsize=8)

    # Panel 3: HIV testing & STI screening
    ax = axes[1, 0]
    hiv_test = [sa_dfs[sc]['hiv_tested_12mo'].mean() * 100 for sc in SCENARIOS if sc in sa_dfs]
    sti_scr = [sa_dfs[sc]['sti_screened'].mean() * 100 for sc in SCENARIOS if sc in sa_dfs]
    ax.bar(x - width/2, hiv_test, width, label='HIV Tested', color='#e74c3c', alpha=0.8)
    ax.bar(x + width/2, sti_scr, width, label='STI Screened', color='#9b59b6', alpha=0.8)
    ax.set_xticks(x)
    ax.set_xticklabels(['Youth Clinic', 'Public Facility', 'Rural'], fontsize=8)
    ax.set_ylabel('Percentage (%)')
    ax.set_title('HIV Testing & STI Screening (Sexually Active)')
    ax.legend(fontsize=8)

    # Panel 4: GBV & depression
    ax = axes[1, 1]
    gbv = []
    dep = []
    for sc in SCENARIOS:
        if sc in dfs:
            d = dfs[sc]
            gbv_s = d[d['gbv_screened'] == 1]
            dep_s = d[d['depression_screened'] == 1]
            gbv.append(gbv_s['gbv_positive'].mean() * 100 if len(gbv_s) else 0)
            dep.append(dep_s['depression_positive'].mean() * 100 if len(dep_s) else 0)
    ax.bar(x - width/2, gbv, width, label='GBV+', color='#e74c3c', alpha=0.8)
    ax.bar(x + width/2, dep, width, label='Depression+', color='#9b59b6', alpha=0.8)
    ax.set_xticks(x)
    ax.set_xticklabels(['Youth Clinic', 'Public Facility', 'Rural'], fontsize=8)
    ax.set_ylabel('Prevalence (%)')
    ax.set_title('GBV & Depression Across Scenarios')
    ax.legend(fontsize=8)

    # Panel 5: Contraceptive method distribution
    ax = axes[2, 0]
    users = df[(df['using_modern_contraceptive'] == 1)]
    if len(users) > 0:
        meth = users['contraceptive_method'].value_counts()
        colors = ['#2ecc71', '#3498db', '#f39c12', '#e74c3c', '#9b59b6', '#1abc9c']
        ax.pie(meth.values, labels=[m.replace('_', ' ').title() for m in meth.index],
               autopct='%1.1f%%', colors=colors[:len(meth)], startangle=90,
               textprops={'fontsize': 7})
    ax.set_title('Contraceptive Method Mix')

    # Panel 6: Age distribution
    ax = axes[2, 1]
    ax.hist(df['age_years'], bins=15, color='#3498db', alpha=0.7, edgecolor='white')
    ax.set_xlabel('Age (years)')
    ax.set_title('Age Distribution')

    # Panel 7: Unmet need across scenarios
    ax = axes[3, 0]
    unmet = [sa_dfs[sc]['unmet_need_contraception'].mean() * 100 for sc in SCENARIOS if sc in sa_dfs]
    ax.bar(x, unmet, color=['#2ecc71', '#f39c12', '#e74c3c'], alpha=0.8)
    ax.set_xticks(x)
    ax.set_xticklabels(['Youth Clinic', 'Public Facility', 'Rural'], fontsize=8)
    for i, v in enumerate(unmet):
        ax.text(i, v + 0.5, f'{v:.1f}%', ha='center', fontsize=10)
    ax.set_ylabel('Percentage (%)')
    ax.set_title('Unmet Need for Contraception')

    # Panel 8: Education level
    ax = axes[3, 1]
    edu = df['education_level'].value_counts()
    edu_order = ['none', 'primary', 'lower_secondary', 'upper_secondary', 'tertiary']
    edu_colors = ['#e74c3c', '#f39c12', '#f1c40f', '#2ecc71', '#3498db']
    vals = [edu.get(e, 0) for e in edu_order]
    ax.bar(range(5), vals, color=edu_colors)
    ax.set_xticks(range(5))
    ax.set_xticklabels(['None', 'Primary', 'Lower Sec', 'Upper Sec', 'Tertiary'], fontsize=7)
    ax.set_ylabel('Count')
    ax.set_title('Education Level')

    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)