training_encoders.py

import pandas as pd
from category_encoders import TargetEncoder
import joblib

# Load your training data
df = pd.read_excel("dataset/Claim Fraud Detection.xlsx")   # change to your actual filename

categorical_cols = [
    'auto_model', 'auto_make', 'police_report_available', 'property_damage',
    'incident_city', 'incident_state', 'authorities_contacted',
    'incident_severity', 'collision_type', 'incident_type',
    'insured_relationship', 'insured_hobbies', 'insured_occupation',
    'insured_education_level', 'insured_sex', 'policy_csl', 'policy_state'
]

target_col = 'fraud_reported'

encoders = {}

for col in categorical_cols:
    te = TargetEncoder()
    te.fit(df[col], df[target_col])
    encoders[col] = te   # store encoder per column

# ✅ Save the dict of encoders
joblib.dump(encoders, "encoders.pkl")
print("Saved encoders as dictionary with keys:", list(encoders.keys()))


encoders = joblib.load("encoders.pkl")
print(type(encoders))
print(list(encoders.keys()))