validate_dataset.py

#!/usr/bin/env python3
"""Validation for AI & Clinical Decision Support Dataset."""
import pandas as pd, numpy as np, matplotlib.pyplot as plt, os, glob

def load_scenarios(data_dir='data'):
    dfs = {}
    for f in sorted(glob.glob(os.path.join(data_dir, 'ai_cdss_*.csv'))):
        name = os.path.basename(f).replace('.csv', '')[8:]
        dfs[name] = pd.read_csv(f)
    return dfs

def main():
    dfs = load_scenarios()
    if not dfs: return
    all_df = pd.concat([df.assign(scenario=n) for n, df in dfs.items()], ignore_index=True)
    fig, axes = plt.subplots(4, 2, figsize=(16, 20))
    fig.suptitle('AI & Clinical Decision Support — Validation Report', fontsize=14, fontweight='bold', y=0.98)
    colors = {'ai_deployed': '#2ecc71', 'ai_pilot': '#f39c12', 'ai_absent': '#e74c3c'}
    labels = {'ai_deployed': 'Deployed (SA/KE)', 'ai_pilot': 'Pilot (KE/GH)', 'ai_absent': 'Absent (DRC/SLE)'}
    scenarios = list(dfs.keys())

    ax = axes[0, 0]
    metrics = ['AI Avail %', 'CDSS %', 'AI Used %', 'Rec Followed %', 'Validated %']
    for i, s in enumerate(scenarios):
        d = dfs[s]; vals = [d['ai_tool_available'].mean()*100, d['cdss_available'].mean()*100, d['ai_used_this_encounter'].mean()*100, d['recommendation_followed'].mean()*100, d['validated_local_population'].mean()*100]
        ax.bar(np.arange(len(metrics))+i*0.25, vals, 0.25, label=labels.get(s,s), color=colors[s], alpha=0.8)
    ax.set_xticks(np.arange(len(metrics))+0.25); ax.set_xticklabels(metrics, fontsize=6); ax.set_ylabel('%'); ax.set_title('Panel 1: Key Metrics'); ax.legend(fontsize=6)

    ax = axes[0, 1]
    domains = ['tb_xray_screening','diabetic_retinopathy','triage_severity','maternal_risk','malaria_microscopy','clinical_guidelines']
    for i, s in enumerate(scenarios):
        vals = [dfs[s]['ai_domain'].value_counts(normalize=True).get(d,0)*100 for d in domains]
        ax.bar(np.arange(len(domains))+i*0.20, vals, 0.20, label=labels.get(s,s), color=colors[s], alpha=0.8)
    ax.set_xticks(np.arange(len(domains))+0.20); ax.set_xticklabels([d.replace('_','\n') for d in domains], fontsize=3); ax.set_ylabel('%'); ax.set_title('Panel 2: AI Domain'); ax.legend(fontsize=6)

    ax = axes[1, 0]
    types = ['image_recognition','rule_based','predictive_model','hybrid','nlp']
    for i, s in enumerate(scenarios):
        vals = [dfs[s]['ai_type'].value_counts(normalize=True).get(t,0)*100 for t in types]
        ax.bar(np.arange(len(types))+i*0.25, vals, 0.25, label=labels.get(s,s), color=colors[s], alpha=0.8)
    ax.set_xticks(np.arange(len(types))+0.25); ax.set_xticklabels([t.replace('_','\n') for t in types], fontsize=5); ax.set_ylabel('%'); ax.set_title('Panel 3: AI Type'); ax.legend(fontsize=6)

    ax = axes[1, 1]
    modes = ['cloud','edge_device','offline','hybrid']
    for i, s in enumerate(scenarios):
        vals = [dfs[s]['deployment_mode'].value_counts(normalize=True).get(m,0)*100 for m in modes]
        ax.bar(np.arange(len(modes))+i*0.25, vals, 0.25, label=labels.get(s,s), color=colors[s], alpha=0.8)
    ax.set_xticks(np.arange(len(modes))+0.25); ax.set_xticklabels(modes); ax.set_ylabel('%'); ax.set_title('Panel 4: Deployment Mode'); ax.legend(fontsize=7)

    ax = axes[2, 0]
    barriers = ['Infra', 'Trust', 'Training', 'Regulation', 'Data Quality', 'Cost', 'Bias']
    for i, s in enumerate(scenarios):
        d = dfs[s]; vals = [d['barrier_infrastructure'].mean()*100, d['barrier_trust'].mean()*100, d['barrier_training'].mean()*100, d['barrier_regulation'].mean()*100, d['barrier_data_quality'].mean()*100, d['barrier_cost'].mean()*100, d['barrier_bias_concern'].mean()*100]
        ax.bar(np.arange(len(barriers))+i*0.20, vals, 0.20, label=labels.get(s,s), color=colors[s], alpha=0.8)
    ax.set_xticks(np.arange(len(barriers))+0.20); ax.set_xticklabels(barriers, fontsize=5); ax.set_ylabel('%'); ax.set_title('Panel 5: Barriers'); ax.legend(fontsize=6)

    ax = axes[2, 1]
    out = ['Accuracy\nImproved', 'Outcome\nImproved', 'False Pos', 'False Neg', 'Regulatory\nApproved', 'Open Source']
    for i, s in enumerate(scenarios):
        d = dfs[s]; vals = [d['diagnostic_accuracy_improved'].mean()*100, d['patient_outcome_improved'].mean()*100, d['false_positive'].mean()*100, d['false_negative'].mean()*100, d['regulatory_approved'].mean()*100, d['open_source'].mean()*100]
        ax.bar(np.arange(len(out))+i*0.20, vals, 0.20, label=labels.get(s,s), color=colors[s], alpha=0.8)
    ax.set_xticks(np.arange(len(out))+0.20); ax.set_xticklabels(out, fontsize=4); ax.set_ylabel('%'); ax.set_title('Panel 6: Outcomes & Safety'); ax.legend(fontsize=6)

    ax = axes[3, 0]
    for s in scenarios:
        ax.hist(dfs[s]['provider_trust_ai'], bins=10, alpha=0.5, label=labels.get(s,s), color=colors[s], density=True)
    ax.set_xlabel('Trust Score (1-10)'); ax.set_title('Panel 7: Provider Trust'); ax.legend(fontsize=7)

    ax = axes[3, 1]
    nc = ['ai_tool_available','ai_used_this_encounter','recommendation_followed','diagnostic_accuracy_improved','provider_trust_ai','internet_available']
    corr = all_df[nc].corr()
    im = ax.imshow(corr, cmap='RdBu_r', vmin=-1, vmax=1, aspect='auto')
    ax.set_xticks(range(len(nc))); ax.set_yticks(range(len(nc)))
    ax.set_xticklabels([c.replace('_','\n') for c in nc], fontsize=5, rotation=45, ha='right')
    ax.set_yticklabels([c.replace('_','\n') for c in nc], fontsize=5)
    ax.set_title('Panel 8: Correlations'); fig.colorbar(im, ax=ax, fraction=0.046)

    plt.tight_layout(rect=[0,0,1,0.96]); plt.savefig('validation_report.png', dpi=150, bbox_inches='tight'); plt.close()
    print("Saved validation_report.png")

if __name__ == '__main__':
    main()