utils/data_processing.py

import pandas as pd

def convert_dates(df, columns):
    for col in columns:
        if col in df.columns:
            df[col] = pd.to_datetime(df[col], errors='coerce')
    return df

def calculate_age(df, dob_col="DOB"):
    if dob_col in df.columns:
        df["Age"] = (pd.to_datetime("today") - df[dob_col]).dt.days // 365
    return df

def clean_performance_score(df, col="Performance Score"):
    score_map = {"Exceeds": 5, "Fully Meets": 4, "Needs Improvement": 3, "PIP": 2}
    if col in df.columns:
        df[col] = df[col].map(lambda x: score_map.get(str(x).strip(), None))
        df[col] = pd.to_numeric(df[col], errors='coerce')
    return df