고성능 데이터 파이프라인을 위한 강력한 Python 데코레이터 5가지

from numba import njit
import numpy as np

@njit
def compute_complex_metric(income_array):
    result = np.zeros_like(income_array)
    for i in range(len(income_array)):
        result[i] = np.log1p(income_array[i] * 2.5) ** 1.5
    return result

from joblib import Memory

memory = Memory(".pipeline_cache", verbose=0)

@memory.cache
def expensive_aggregation(df):
    # 중복 계산을 피하기 위해 결과를 디스크에 캐싱
    return df.groupby('ocean_proximity', as_index=False).mean(numeric_only=True)

import pandera as pa

housing_schema = pa.DataFrameSchema({
    "median_income": pa.Column(float, pa.Check.greater_than(0)),
    "total_rooms": pa.Column(float, pa.Check.gt(0))
})

@pa.check_types
def validate_and_process(df: pd.DataFrame) -> pa.typing.DataFrame:
    return housing_schema.validate(df)