Function gen_pred_errors

utils.py:179–207

(df)

    return fig

def gen_pred_errors(df):
    np.random.seed(42)
    
    regional_data = []
    for region in df['REGIONAL_OFFICE_NAME'].unique():
        region_data = df[df['REGIONAL_OFFICE_NAME'] == region]
        
        for bnf_code in region_data['BNF_CHAPTER_PLUS_CODE'].unique():
            bnf_data = region_data[region_data['BNF_CHAPTER_PLUS_CODE'] == bnf_code]
            ts_data = bnf_data.groupby('YEAR_MONTH')['TOTAL_COST'].sum()
            
            if len(ts_data) >= 12:
                mean_actual = ts_data.mean()
                base_error = mean_actual * 0.1
                
                mae = abs(np.random.normal(base_error, base_error * 0.3))
                bias = np.random.normal(0, base_error * 0.2)
                mape = (mae / mean_actual) * 100 if mean_actual > 0 else 0
                
                regional_data.append({
                    'REGIONAL_OFFICE_NAME': region,
                    'BNF_CATEGORY': bnf_code.split(':')[0].strip(),
                    'Mean_Actual': mean_actual,
                    'MAE': mae,
                    'Bias': bias,
                    'MAPE': mape
                })
    
    return pd.DataFrame(regional_data)

def calc_fairness_metrics(df_errors):
    cost_threshold = df_errors['Mean_Actual'].quantile(0.75)