筛选相关特征

def calcute_corr(data,aim):
    '''
    data : 包含所有需要计算corr相关性的特征数据集
    aim : 相关性删除的阈值
    return :
        result : 留下待使用的特征
        the_same_feature : 要删除的相关性大于阈值的特征
    '''
    the_same_feature = []
    result = []
    for col_name in data.columns:
        if col_name in the_same_feature:
            continue
        else:
            temp = data.corrwith(data[col_name])
            result.append(col_name)
            te = list(temp[temp > aim].index)
            the_same_feature.extend(te)
    return result,the_same_feature