class CustConfig: FEATURE_COLUMNS = [ "BB_RETAIL_CUSTOMER_CODE", # 零售户代码 "BB_RTL_CUST_MARKET_TYPE_NAME", # 零售户市场类型名称 "BB_RTL_CUST_BUSINESS_TYPE_NAME", # 零售客户业态名称 "BB_RTL_CUST_CHAIN_FLAG", # 零售户连锁标识 "MD04_MG_RTL_CUST_CREDITCLASS_NAME", # 零售户信用等级名称 "MD04_DIR_SAL_STORE_FLAG", # 直营店标识 "BB_CUSTOMER_MANAGER_SCOPE_NAME", # 零售户经营范围名称 "BB_RTL_CUST_TERMINAL_LEVEL_NAME", # 零售户终端层级名称 "OPERATOR_EDU", # 零售客户经营者文化程度 "STORE_AREA", # 店铺经营面积 "OPERATOR_AGE", # 经营者年龄 "PRODUCT_INSALE_QTY", # 在销品规数 ] ONEHOT_CAT = { "BB_RTL_CUST_MARKET_TYPE_NAME": ["城网", "农网"], "BB_RTL_CUST_BUSINESS_TYPE_NAME": ["便利店", "超市", "烟草专业店", "娱乐服务类", "其他"], "BB_RTL_CUST_CHAIN_FLAG": ["是", "否"], "MD04_MG_RTL_CUST_CREDITCLASS_NAME": ["AAA", "AA", "A", "B", "C", "D"], "MD04_DIR_SAL_STORE_FLAG": ["是", "否"], "BB_CUSTOMER_MANAGER_SCOPE_NAME": ["是", "否"], "BB_RTL_CUST_TERMINAL_LEVEL_NAME": ["普通终端", "一般现代终端", "合作终端", "加盟终端", "直营终端"], "OPERATOR_EDU": [1, 2, 3, 4, 5, 6, 7, "无数据"], "STORE_AREA": ["0-20", "20-30", "30-40", "40-60", "60-80", "80-100", "100-120", "120以上"], "OPERATOR_AGE": ["18-30", "31-40", "41-50", "51-65", "66-80", "80以上"], "PRODUCT_INSALE_QTY": ["0-10", "11-20", "21-30", "31-40", "41-50", "51-60", "61-70", "71-80", "81-90", "91-100", "101-110", "111-120", "121-130", "131-140", "141-150", "151-160", "161-170", "171-180", "181-190", "191-200", "201-210", "211-220", "221-230", "231-240", "241-250", "251-260", "261-270", "271-280", "281-290", "291-350"], } CLEANING_RULES = { "BB_RTL_CUST_MARKET_TYPE_NAME": {"method": "fillna", "opt": "fill", "value": "城网", "type": "str"}, "BB_RTL_CUST_BUSINESS_TYPE_NAME": {"method": "fillna", "opt": "fill", "value": "其他", "type": "str"}, "BB_RTL_CUST_CHAIN_FLAG": {"method": "fillna", "opt": "fill", "value": "否", "type": "str"}, "MD04_MG_RTL_CUST_CREDITCLASS_NAME": {"method": "fillna", "opt": "fill", "value": "B", "type": "str"}, "MD04_DIR_SAL_STORE_FLAG": {"method": "fillna", "opt": "fill", "value": "否", "type": "str"}, "BB_CUSTOMER_MANAGER_SCOPE_NAME": {"method": "fillna", "opt": "fill", "value": "否", "type": "str"}, "BB_RTL_CUST_TERMINAL_LEVEL_NAME": {"method": "fillna", "opt": "fill", "value": "普通终端", "type": "str"}, "OPERATOR_EDU": {"method": "fillna", "opt": "fill", "value": "无数据", "type": "str"}, "STORE_AREA": {"method": "fillna", "opt": "fill", "value": "0-20", "type": "str"}, "OPERATOR_AGE": {"method": "fillna", "opt": "fill", "value": "31-40", "type": "str"}, "PRODUCT_INSALE_QTY": {"method": "fillna", "opt": "fill", "value": "0-10", "type": "str"}, } class ProductConfig: FEATURE_COLUMNS = [ "product_code", # 商品编码 "factory_name", # 产地(工业公司名称) "brand_name", # 品牌名称 "is_low_tar", # 低焦油卷烟 "is_medium", # 中支烟 "is_tiny", # 细支烟 "is_coarse", # 粗支烟（同时非中非细） "is_exploding_beads", # 爆珠烟 "is_abnormity", # 异形包装 "is_cig", # 雪茄烟 "is_chuangxin", # 创新品类 "direct_retail_price", # 卷烟建议零售价 "tbc_total_length", # 烟支总长度 "product_style", # 包装类型 ] ONEHOT_CAT = { "factory_name": ["安徽中烟", "澳门云福卷烟厂", "北欧烟草集团", "博格集团", "重庆中烟", "川渝中烟", "菲利普莫里斯亚洲", "福建中烟", "甘肃工业", "广东中烟", "广西中烟", "贵州中烟", "海南红塔", "河北中烟", "河南中烟", "黑龙江工业", "红塔辽宁烟草", "湖北中烟", "湖南中烟", "吉林工业", "家源开发股份有限公司", "嘉莱赫国际有限公司", "江苏中烟", "江西中烟", "凯德控股有限公司", "力量雪茄烟草有限公司", "南洋兄弟烟草股份", "内蒙古昆明卷烟", "日本烟草（香港）有限公司", "三宝麟国际集团", "厦门调拨站", "山东中烟", "山西昆明烟草", "陕西中烟", "上海烟草(集团)公司", "上海烟草公司", "深圳工业", "四川中烟", "特富意烟草(国际)", "雪茄客烟草国际贸易有限公司", "耀莱雪茄控股有限公司", "引领国际有限公司", "英飞烽香港有限公司", "英美烟草中国有限公司", "云南中烟", "浙江中烟", "中茄国际贸易有限公司", "中烟英美烟草国际有限公司", "株式会社 KT&G", "无"], "brand_name": ["万宝路", "555", "骆驼(国外)", "大华", "娇子", "大青山", "龙凤呈祥", "黄鹤楼", "真龙", "七匹狼", "芙蓉王", "双喜(广)", "贵烟", "钓鱼台", "红双喜(南洋)", "云烟", "蒙特", "富恩特", "拉·加莱拉", "苏烟", "丹纳曼", "黄山", "南京", "利群", "金桥", "泰山", "好日子", "石林", "美登", "红河", "嘉辉", "七星", "都彭", "天下秀", "长城", "高希霸", "钻石", "金圣", "王冠雪茄", "黄金叶", "中南海", "长白山", "红旗渠", "建牌", "大卫杜夫", "罗密欧", "茂大", "红金龙", "天子", "熊猫", "双喜(深)", "大前门", "兰州", "红双喜(沪)", "雄狮", "广州", "红玫王", "黄果树", "红塔山", "福", "小熊猫", "爱喜", "蒙特利", "玉溪", "都宝", "麦克纽杜", "卡里罗", "中华", "牡丹(沪)", "阿里山", "顺百利", "白沙", "羊城", "白云", "特美思", "国宾", "帕特加", "比德奥", "冬虫夏草", "威龙（湛江）", "香格里拉", "红梅", "延安", "特富意", "石狮", "金香港", "好猫", "登喜路", "乐迪", "林海灵芝", "椰树", "北京", "大红鹰", "大丰收", "红双喜(武汉)", "五叶神", "狮", "优民", "将军", "遵义", "恒大", "飞马", "红三环", "芙蓉", "工字", "古田", "狮牌", "君力", "哈尔滨", "梦都", "香梅(阜阳)", "哈德门", "梅州", "红山茶", "猴王", "沙龙", "潘趣", "狮子牌", "上海", "红玫", "醒宝", "广州湾", "百乐门", "关塔那摩", "威斯", "五一", "寿百年", "人民大会堂", "土楼", "三沙", "西湖", "光明", "阿诗玛", "宝亨", "恭贺新禧", "长寿", "茶花", "迎客松", "龙烟", "金澳门", "宝岛", "多米尼加之花", "国喜", "金驼", "君特欧", "上游", "幸福", "春城", "吉庆", "黄山松", "黄金龙", "紫气东来", "彼亚赛", "银辉", "潮牌", "庐山", "三峡", "壹支笔", "双叶", "无"], "is_low_tar": ["是", "否"], "is_medium": ["是", "否"], "is_tiny": ["是", "否"], "is_coarse": ["是", "否"], "is_exploding_beads": ["是", "否"], "is_abnormity": ["是", "否"], "is_cig": ["是", "否"], "is_chuangxin": ["是", "否"], "direct_retail_price": ["0-5", "5-10", "10-15", "15-20", "20-26", "26-30", "30-40", "40-50", "50-65", "65-80", "80-100", "100以上", "5-9.9", "10-19.9", "20-29.9", "30-39.9", "40-49.9", "50-59.9", "60-69.9", "70-79.9", "80-89.9", "90-99.9", "100-109.9", "110-119.9", "120-129.9", "130-139.9", "140-149.9", "150-199.9", "200-249.9", "250-499.9", "500以上"], "tbc_total_length": ["小于79", "80-88", "89-100", "大于120"], "product_style": ["条盒硬盒", "条包硬盒", "条盒软盒", "条包软盒", "铁盒", "其他"], } CLEANING_RULES = { "factory_name": {"method": "fillna", "opt": "fill", "value": "无", "type": "str"}, "brand_name": {"method": "fillna", "opt": "fill", "value": "无", "type": "str"}, "is_low_tar": {"method": "fillna", "opt": "fill", "value": "否", "type": "str"}, "is_medium": {"method": "fillna", "opt": "fill", "value": "否", "type": "str"}, "is_tiny": {"method": "fillna", "opt": "fill", "value": "否", "type": "str"}, "is_coarse": {"method": "fillna", "opt": "fill", "value": "否", "type": "str"}, "is_exploding_beads": {"method": "fillna", "opt": "fill", "value": "否", "type": "str"}, "is_abnormity": {"method": "fillna", "opt": "fill", "value": "否", "type": "str"}, "is_cig": {"method": "fillna", "opt": "fill", "value": "否", "type": "str"}, "is_chuangxin": {"method": "fillna", "opt": "fill", "value": "否", "type": "str"}, "direct_retail_price": {"method": "fillna", "opt": "fill", "value": "0-5", "type": "str"}, "tbc_total_length": {"method": "fillna", "opt": "fill", "value": "小于79", "type": "str"}, "product_style": {"method": "fillna", "opt": "fill", "value": "其他", "type": "str"}, } class OrderConfig: FEATURE_COLUMNS = [ "cust_code", # 零售户编码 "product_code", # 品牌规格编码 "sale_qty", # 销量包 # "sale_qty_l", # 销量上期 # "sale_qty_hb", # 销量环比 # "sale_amt", # 销售额包 ] class ShopConfig: FEATURE_COLUMNS = [ "cust_code", # 客户编码 "r_home_num", # 常驻人口_居住人数 "r_work_num", # 常驻人口_工作人数 "r_resident_num", # 常驻人口_工作或居住人数 "r_urban_cons_middle", # 常驻人口_城市消费水平_中 "r_urban_cons_low", # 常驻人口_城市消费水平_低 "r_urban_cons_lower", # 常驻人口_城市消费水平_次低 "r_urban_cons_secondhigh", # 常驻人口_城市消费水平_次高 "r_urban_cons_high", # 常驻人口_城市消费水平_高 "r_edu_junior_middle", # 常驻人口_学历_初中 "r_edu_doctor", # 常驻人口_学历_博士 "r_edu_specialty", # 常驻人口_学历_大专 "r_edu_primary", # 常驻人口_学历_小学 "r_edu_college", # 常驻人口_学历_本科 "r_edu_postgraduate", # 常驻人口_学历_硕士 "r_edu_senior_middle", # 常驻人口_学历_高中 "r_house_price79999", # 常驻人口_居住社区房价_60000_79999 "r_house_price59999", # 常驻人口_居住社区房价_40000_59999 "r_house_price39999", # 常驻人口_居住社区房价_20000_39999 "r_house_price19999", # 常驻人口_居住社区房价_10000_19999 "r_house_price9999", # 常驻人口_居住社区房价_8000_9999 "r_house_price7999", # 常驻人口_居住社区房价_5000_7999 "r_house_price4999", # 常驻人口_居住社区房价_2000_4999 "r_age_17", # 常驻人口_年龄_0_17 "r_age_24", # 常驻人口_年龄_18_24 "r_age_30", # 常驻人口_年龄_25_30 "r_age_35", # 常驻人口_年龄_31_35 "r_age_40", # 常驻人口_年龄_36_40 "r_age_45", # 常驻人口_年龄_41_45 "r_age_60", # 常驻人口_年龄_46_60 "r_age_over_60", # 常驻人口_年龄_61以上 "r_sex_woman", # 常驻人口_性别_女 "r_sex_man", # 常驻人口_性别_男 "r_catering_50", # 常驻人口_餐饮消费水平_50 "r_catering_100", # 常驻人口_餐饮消费水平_100 "r_catering_150", # 常驻人口_餐饮消费水平_150 "r_catering_200", # 常驻人口_餐饮消费水平_200 "r_catering_500", # 常驻人口_餐饮消费水平_500 "r_catering_over_500", # 常驻人口_餐饮消费水平_500以上 "r_catering_times_2", # 常驻人口_餐饮消费频次_1_2 "r_catering_times_4", # 常驻人口_餐饮消费频次_2_4 "r_catering_times_6", # 常驻人口_餐饮消费频次_4_6 "r_catering_times_8", # 常驻人口_餐饮消费频次_6_8 "r_catering_times_10", # 常驻人口_餐饮消费频次_8_10 "r_catering_times_11", # 常驻人口_餐饮消费频次_11以上 "r_native_beijing", # 常驻人口_家乡地_北京市 "r_native_tianjing", # 常驻人口_家乡地_天津市 "r_native_hebei", # 常驻人口_家乡地_河北省 "r_native_shanxi", # 常驻人口_家乡地_山西省 "r_native_neimeng", # 常驻人口_家乡地_内蒙古 "r_native_liaoning", # 常驻人口_家乡地_辽宁省 "r_native_jilin", # 常驻人口_家乡地_吉林省 "r_native_heilongjiang", # 常驻人口_家乡地_黑龙江省 "r_native_shanghai", # 常驻人口_家乡地_上海市 "r_native_jiangsu", # 常驻人口_家乡地_江苏省 "r_native_zhejiang", # 常驻人口_家乡地_浙江省 "r_native_anhui", # 常驻人口_家乡地_安徽省 "r_native_fujian", # 常驻人口_家乡地_福建省 "r_native_jiangix", # 常驻人口_家乡地_江西省 "r_native_shandong", # 常驻人口_家乡地_山东省 "r_native_henan", # 常驻人口_家乡地_河南省 "r_native_hubei", # 常驻人口_家乡地_湖北省 "r_native_hunan", # 常驻人口_家乡地_湖南省 "r_native_guangdong", # 常驻人口_家乡地_广东省 "r_native_hainan", # 常驻人口_家乡地_海南省 "r_native_sichuan", # 常驻人口_家乡地_四川省 "r_native_guizhou", # 常驻人口_家乡地_贵州省 "r_native_yunnan", # 常驻人口_家乡地_云南省 "r_native_shan", # 常驻人口_家乡地_陕西省 "r_native_gansu", # 常驻人口_家乡地_甘肃省 "r_native_qinghai", # 常驻人口_家乡地_青海省 "r_native_guangxi", # 常驻人口_家乡地_广西壮族自治区 "r_native_ningxia", # 常驻人口_家乡地_宁夏回族自治区 "r_native_xinjiang", # 常驻人口_家乡地_新疆维吾尔自治区 "r_native_xizang", # 常驻人口_家乡地_西藏自治区 "r_native_chongqing", # 常驻人口_家乡地_重庆市 "r_native_hongkong", # 常驻人口_家乡地_香港 "r_native_macao", # 常驻人口_家乡地_澳门 "r_native_taiwan", # 常驻人口_家乡地_台湾 "r_native_other", # 常驻人口_家乡地_其它 "f_flow_num", # 流动人口_工作日_日均流动人口数量 "f_holiday_flow_num", # 流动人口_节假日_日均流动人口数量 "f_workday_flow_num", # 流动人口_日均流动人口数量 "f_flowurban_cons_middle", # 日均流动_城市消费水平_中 "f_flowurban_cons_low", # 日均流动_城市消费水平_低 "f_flowurban_cons_lower", # 日均流动_城市消费水平_次低 "f_flowurban_cons_second_high", # 日均流动_城市消费水平_次高 "f_flowurban_cons_high", # 日均流动_城市消费水平_高 "f_flowedu_junior_middle", # 日均流动_学历_初中 "f_flowedu_doctor", # 日均流动_学历_博士 "f_flowedu_specialty", # 日均流动_学历_大专 "f_flowedu_primary", # 日均流动_学历_小学 "f_flowedu_college", # 日均流动_学历_本科 "f_flowedu_postgraduate", # 日均流动_学历_硕士 "f_flowedu_senior_middle", # 日均流动_学历_高中 "f_flowhouse_middle", # 日均流动_居住社区房价_中 "f_flowhouse_low", # 日均流动_居住社区房价_低 "f_flowhouse_lower", # 日均流动_居住社区房价_次低 "f_flowhouse_second_high", # 日均流动_居住社区房价_次高 "f_flowhouse_high", # 日均流动_居住社区房价_高 "f_flowage_17", # 日均流动_年龄_0_17 "f_flowage_24", # 日均流动_年龄_18_24 "f_flowage_30", # 日均流动_年龄_25_30 "f_flowage_35", # 日均流动_年龄_31_35 "f_flowage_40", # 日均流动_年龄_36_40 "f_flowage_45", # 日均流动_年龄_41_45 "f_flowage_60", # 日均流动_年龄_46_60 "f_flowage_over_60", # 日均流动_年龄_61以上 "f_flowsex_woman", # 日均流动_性别_女 "f_flowsex_man", # 日均流动_性别_男 "f_holidayurban_cons_middle", # 节假日流动_城市消费水平_中 "f_holidayurban_cons_low", # 节假日流动_城市消费水平_低 "f_holidayurban_cons_lower", # 节假日流动_城市消费水平_次低 "f_holidayurban_cons_secondhigh", # 节假日流动_城市消费水平_次高 "f_holidayurban_cons_high", # 节假日流动_城市消费水平_高 "f_holidayedu_junior_middle", # 节假日流动_学历_初中 "f_holidayedu_doctor", # 节假日流动_学历_博士 "f_holidayedu_specialty", # 节假日流动_学历_大专 "f_holidayedu_primary", # 节假日流动_学历_小学 "f_holidayedu_college", # 节假日流动_学历_本科 "f_holidayedu_postgraduate", # 节假日流动_学历_硕士 "f_holidayedu_senior_middle", # 节假日流动_学历_高中 "f_holidayhouse_middle", # 节假日流动_居住社区房价_中 "f_holidayhouse_low", # 节假日流动_居住社区房价_低 "f_holidayhouse_lower", # 节假日流动_居住社区房价_次低 "f_holidayhouse_second_high", # 节假日流动_居住社区房价_次高 "f_holidayhouse_high", # 节假日流动_居住社区房价_高 "f_holidayage_17", # 节假日流动_年龄_0_17 "f_holidayage_24", # 节假日流动_年龄_18_24 "f_holidayage_30", # 节假日流动_年龄_25_30 "f_holidayage_35", # 节假日流动_年龄_31_35 "f_holidayage_40", # 节假日流动_年龄_36_40 "f_holidayage_45", # 节假日流动_年龄_41_45 "f_holidayage_60", # 节假日流动_年龄_46_60 "f_holidayage_over_60", # 节假日流动_年龄_61以上 "f_holidaysex_woman", # 节假日流动_性别_女 "f_holidaysex_man", # 节假日流动_性别_男 "f_workday_urban_cons_middle", # 工作日流动_城市消费水平_中 "f_workday_urban_cons_low", # 工作日流动_城市消费水平_低 "f_workday_urban_cons_lower", # 工作日流动_城市消费水平_次低 "f_workday_urban_cons_secondhigh",# 工作日流动_城市消费水平_次高 "f_workday_urban_cons_high", # 工作日流动_城市消费水平_高 "f_workday_edu_junior_middle", # 工作日流动_学历_初中 "f_workday_edu_doctor", # 工作日流动_学历_博士 "f_workday_edu_specialty", # 工作日流动_学历_大专 "f_workday_edu_primary", # 工作日流动_学历_小学 "f_workday_edu_college", # 工作日流动_学历_本科 "f_workday_edu_postgraduate", # 工作日流动_学历_硕士 "f_workday_edu_senior_middle", # 工作日流动_学历_高中 "f_workday_house_middle", # 工作日流动_居住社区房价_中 "f_workday_house_low", # 工作日流动_居住社区房价_低 "f_workday_house_lower", # 工作日流动_居住社区房价_次低 "f_workday_house_second_high", # 工作日流动_居住社区房价_次高 "f_workday_house_high", # 工作日流动_居住社区房价_高 "f_workday_age_17", # 工作日流动_年龄_0_17 "f_workday_age_24", # 工作日流动_年龄_18_24 "f_workday_age_30", # 工作日流动_年龄_25_30 "f_workday_age_35", # 工作日流动_年龄_31_35 "f_workday_age_40", # 工作日流动_年龄_36_40 "f_workday_age_45", # 工作日流动_年龄_41_45 "f_workday_age_60", # 工作日流动_年龄_46_60 "f_workday_age_over_60", # 工作日流动_年龄_61以上 "f_workday_sex_woman", # 工作日流动_性别_女 "f_workday_sex_man", # 工作日流动_性别_男 ] ONEHOT_CAT = { "r_home_num": ["0-100", "101-500", "501-2000", "2001-5000", "5001-10000", "10000以上"], "r_work_num": ["0-100", "101-500", "501-2000", "2001-5000", "5001-10000", "10000以上"], "r_resident_num": ["0-100", "101-500", "501-2000", "2001-5000", "5001-10000", "10001-20000", "20000以上"], "r_urban_cons_middle": ["0-10", "10-20", "20-30", "30-40", "40-50", "50-60", "60-70", "70-80", "80-90", "90-100"], "r_urban_cons_low": ["0-10", "10-20", "20-30", "30-40", "40-50", "50-60", "60-70", "70-80", "80-90", "90-100"], "r_urban_cons_lower": ["0-10", "10-20", "20-30", "30-40", "40-50", "50-60", "60-70", "70-80", "80-90", "90-100"], "r_urban_cons_secondhigh": ["0-10", "10-20", "20-30", "30-40", "40-50", "50-60", "60-70", "70-80", "80-90", "90-100"], "r_urban_cons_high": ["0-10", "10-20", "20-30", "30-40", "40-50", "50-60", "60-70", "70-80", "80-90", "90-100"], "r_edu_junior_middle": ["0-10", "10-20", "20-30", "30-40", "40-50", "50-60", "60-70", "70-80", "80-90", "90-100"], "r_edu_doctor": ["0-10", "10-20", "20-30", "30-40", "40-50", "50-60", "60-70", "70-80", "80-90", "90-100"], "r_edu_specialty": ["0-10", "10-20", "20-30", "30-40", "40-50", "50-60", "60-70", "70-80", "80-90", "90-100"], "r_edu_primary": ["0-10", "10-20", "20-30", "30-40", "40-50", "50-60", "60-70", "70-80", "80-90", "90-100"], "r_edu_college": ["0-10", "10-20", "20-30", "30-40", "40-50", "50-60", "60-70", "70-80", "80-90", "90-100"], "r_edu_postgraduate": ["0-10", "10-20", "20-30", "30-40", "40-50", "50-60", "60-70", "70-80", "80-90", "90-100"], "r_edu_senior_middle": ["0-10", "10-20", "20-30", "30-40", "40-50", "50-60", "60-70", "70-80", "80-90", "90-100"], "r_house_price79999": ["0-10", "10-20", "20-30", "30-40", "40-50", "50-60", "60-70", "70-80", "80-90", "90-100"], "r_house_price59999": ["0-10", "10-20", "20-30", "30-40", "40-50", "50-60", "60-70", "70-80", "80-90", "90-100"], "r_house_price39999": ["0-10", "10-20", "20-30", "30-40", "40-50", "50-60", "60-70", "70-80", "80-90", "90-100"], "r_house_price19999": ["0-10", "10-20", "20-30", "30-40", "40-50", "50-60", "60-70", "70-80", "80-90", "90-100"], "r_house_price9999": ["0-10", "10-20", "20-30", "30-40", "40-50", "50-60", "60-70", "70-80", "80-90", "90-100"], "r_house_price7999": ["0-10", "10-20", "20-30", "30-40", "40-50", "50-60", "60-70", "70-80", "80-90", "90-100"], "r_house_price4999": ["0-10", "10-20", "20-30", "30-40", "40-50", "50-60", "60-70", "70-80", "80-90", "90-100"], "r_age_17": ["0-10", "10-20", "20-30", "30-40", "40-50", "50-60", "60-70", "70-80", "80-90", "90-100"], "r_age_24": ["0-10", "10-20", "20-30", "30-40", "40-50", "50-60", "60-70", "70-80", "80-90", "90-100"], "r_age_30": ["0-10", "10-20", "20-30", "30-40", "40-50", "50-60", "60-70", "70-80", "80-90", "90-100"], "r_age_35": ["0-10", "10-20", "20-30", "30-40", "40-50", "50-60", "60-70", "70-80", "80-90", "90-100"], "r_age_40": ["0-10", "10-20", "20-30", "30-40", "40-50", "50-60", "60-70", "70-80", "80-90", "90-100"], "r_age_45": ["0-10", "10-20", "20-30", "30-40", "40-50", "50-60", "60-70", "70-80", "80-90", "90-100"], "r_age_60": ["0-10", "10-20", "20-30", "30-40", "40-50", "50-60", "60-70", "70-80", "80-90", "90-100"], "r_age_over_60": ["0-10", "10-20", "20-30", "30-40", "40-50", "50-60", "60-70", "70-80", "80-90", "90-100"], "r_sex_woman": ["0-10", "10-20", "20-30", "30-40", "40-50", "50-60", "60-70", "70-80", "80-90", "90-100"], "r_sex_man": ["0-10", "10-20", "20-30", "30-40", "40-50", "50-60", "60-70", "70-80", "80-90", "90-100"], "r_catering_50": ["0-10", "10-20", "20-30", "30-40", "40-50", "50-60", "60-70", "70-80", "80-90", "90-100"], "r_catering_100": ["0-10", "10-20", "20-30", "30-40", "40-50", "50-60", "60-70", "70-80", "80-90", "90-100"], "r_catering_150": ["0-10", "10-20", "20-30", "30-40", "40-50", "50-60", "60-70", "70-80", "80-90", "90-100"], "r_catering_200": ["0-10", "10-20", "20-30", "30-40", "40-50", "50-60", "60-70", "70-80", "80-90", "90-100"], "r_catering_500": ["0-10", "10-20", "20-30", "30-40", "40-50", "50-60", "60-70", "70-80", "80-90", "90-100"], "r_catering_over_500": ["0-10", "10-20", "20-30", "30-40", "40-50", "50-60", "60-70", "70-80", "80-90", "90-100"], "r_catering_times_2": ["0-10", "10-20", "20-30", "30-40", "40-50", "50-60", "60-70", "70-80", "80-90", "90-100"], "r_catering_times_4": ["0-10", "10-20", "20-30", "30-40", "40-50", "50-60", "60-70", "70-80", "80-90", "90-100"], "r_catering_times_6": ["0-10", "10-20", "20-30", "30-40", "40-50", "50-60", "60-70", "70-80", "80-90", "90-100"], "r_catering_times_8": ["0-10", "10-20", "20-30", "30-40", "40-50", "50-60", "60-70", "70-80", "80-90", "90-100"], "r_catering_times_10": ["0-10", "10-20", "20-30", "30-40", "40-50", "50-60", "60-70", "70-80", "80-90", "90-100"], "r_catering_times_11": ["0-10", "10-20", "20-30", "30-40", "40-50", "50-60", "60-70", "70-80", "80-90", "90-100"], "r_native_beijing": ["0-10", "10-20", "20-30", "30-40", "40-50", "50-60", "60-70", "70-80", "80-90", "90-100"], "r_native_tianjing": ["0-10", "10-20", "20-30", "30-40", "40-50", "50-60", "60-70", "70-80", "80-90", "90-100"], "r_native_hebei": ["0-10", "10-20", "20-30", "30-40", "40-50", "50-60", "60-70", "70-80", "80-90", "90-100"], "r_native_shanxi": ["0-10", "10-20", "20-30", "30-40", "40-50", "50-60", "60-70", "70-80", "80-90", "90-100"], "r_native_neimeng": ["0-10", "10-20", "20-30", "30-40", "40-50", "50-60", "60-70", "70-80", "80-90", "90-100"], "r_native_liaoning": ["0-10", "10-20", "20-30", "30-40", "40-50", "50-60", "60-70", "70-80", "80-90", "90-100"], "r_native_jilin": ["0-10", "10-20", "20-30", "30-40", "40-50", "50-60", "60-70", "70-80", "80-90", "90-100"], "r_native_heilongjiang": ["0-10", "10-20", "20-30", "30-40", "40-50", "50-60", "60-70", "70-80", "80-90", "90-100"], "r_native_shanghai": ["0-10", "10-20", "20-30", "30-40", "40-50", "50-60", "60-70", "70-80", "80-90", "90-100"], "r_native_jiangsu": ["0-10", "10-20", "20-30", "30-40", "40-50", "50-60", "60-70", "70-80", "80-90", "90-100"], "r_native_zhejiang": ["0-10", "10-20", "20-30", "30-40", "40-50", "50-60", "60-70", "70-80", "80-90", "90-100"], "r_native_anhui": ["0-10", "10-20", "20-30", "30-40", "40-50", "50-60", "60-70", "70-80", "80-90", "90-100"], "r_native_fujian": ["0-10", "10-20", "20-30", "30-40", "40-50", "50-60", "60-70", "70-80", "80-90", "90-100"], "r_native_jiangix": ["0-10", "10-20", "20-30", "30-40", "40-50", "50-60", "60-70", "70-80", "80-90", "90-100"], "r_native_shandong": ["0-10", "10-20", "20-30", "30-40", "40-50", "50-60", "60-70", "70-80", "80-90", "90-100"], "r_native_henan": ["0-10", "10-20", "20-30", "30-40", "40-50", "50-60", "60-70", "70-80", "80-90", "90-100"], "r_native_hubei": ["0-10", "10-20", "20-30", "30-40", "40-50", "50-60", "60-70", "70-80", "80-90", "90-100"], "r_native_hunan": ["0-10", "10-20", "20-30", "30-40", "40-50", "50-60", "60-70", "70-80", "80-90", "90-100"], "r_native_guangdong": ["0-10", "10-20", "20-30", "30-40", "40-50", "50-60", "60-70", "70-80", "80-90", "90-100"], "r_native_hainan": ["0-10", "10-20", "20-30", "30-40", "40-50", "50-60", "60-70", "70-80", "80-90", "90-100"], "r_native_sichuan": ["0-10", "10-20", "20-30", "30-40", "40-50", "50-60", "60-70", "70-80", "80-90", "90-100"], "r_native_guizhou": ["0-10", "10-20", "20-30", "30-40", "40-50", "50-60", "60-70", "70-80", "80-90", "90-100"], "r_native_yunnan": ["0-10", "10-20", "20-30", "30-40", "40-50", "50-60", "60-70", "70-80", "80-90", "90-100"], "r_native_shan": ["0-10", "10-20", "20-30", "30-40", "40-50", "50-60", "60-70", "70-80", "80-90", "90-100"], "r_native_gansu": ["0-10", "10-20", "20-30", "30-40", "40-50", "50-60", "60-70", "70-80", "80-90", "90-100"], "r_native_qinghai": ["0-10", "10-20", "20-30", "30-40", "40-50", "50-60", "60-70", "70-80", "80-90", "90-100"], "r_native_guangxi": ["0-10", "10-20", "20-30", "30-40", "40-50", "50-60", "60-70", "70-80", "80-90", "90-100"], "r_native_ningxia": ["0-10", "10-20", "20-30", "30-40", "40-50", "50-60", "60-70", "70-80", "80-90", "90-100"], "r_native_xinjiang": ["0-10", "10-20", "20-30", "30-40", "40-50", "50-60", "60-70", "70-80", "80-90", "90-100"], "r_native_xizang": ["0-10", "10-20", "20-30", "30-40", "40-50", "50-60", "60-70", "70-80", "80-90", "90-100"], "r_native_chongqing": ["0-10", "10-20", "20-30", "30-40", "40-50", "50-60", "60-70", "70-80", "80-90", "90-100"], "r_native_hongkong": ["0-10", "10-20", "20-30", "30-40", "40-50", "50-60", "60-70", "70-80", "80-90", "90-100"], "r_native_macao": ["0-10", "10-20", "20-30", "30-40", "40-50", "50-60", "60-70", "70-80", "80-90", "90-100"], "r_native_taiwan": ["0-10", "10-20", "20-30", "30-40", "40-50", "50-60", "60-70", "70-80", "80-90", "90-100"], "r_native_other": ["0-10", "10-20", "20-30", "30-40", "40-50", "50-60", "60-70", "70-80", "80-90", "90-100"], "f_flow_num": ["0-100", "101-500", "501-2000", "2001-5000", "5001-10000", "10001-50000", "50001-100000", "100000以上"], "f_holiday_flow_num": ["0-100", "101-500", "501-2000", "2001-5000", "5001-10000", "10001-50000", "50001-100000", "100000以上"], "f_workday_flow_num": ["0-100", "101-500", "501-2000", "2001-5000", "5001-10000", "10001-50000", "50001-100000", "100000以上"], "f_flowurban_cons_middle": ["0-10", "10-20", "20-30", "30-40", "40-50", "50-60", "60-70", "70-80", "80-90", "90-100"], "f_flowurban_cons_low": ["0-10", "10-20", "20-30", "30-40", "40-50", "50-60", "60-70", "70-80", "80-90", "90-100"], "f_flowurban_cons_lower": ["0-10", "10-20", "20-30", "30-40", "40-50", "50-60", "60-70", "70-80", "80-90", "90-100"], "f_flowurban_cons_second_high": ["0-10", "10-20", "20-30", "30-40", "40-50", "50-60", "60-70", "70-80", "80-90", "90-100"], "f_flowurban_cons_high": ["0-10", "10-20", "20-30", "30-40", "40-50", "50-60", "60-70", "70-80", "80-90", "90-100"], "f_flowedu_junior_middle": ["0-10", "10-20", "20-30", "30-40", "40-50", "50-60", "60-70", "70-80", "80-90", "90-100"], "f_flowedu_doctor": ["0-10", "10-20", "20-30", "30-40", "40-50", "50-60", "60-70", "70-80", "80-90", "90-100"], "f_flowedu_specialty": ["0-10", "10-20", "20-30", "30-40", "40-50", "50-60", "60-70", "70-80", "80-90", "90-100"], "f_flowedu_primary": ["0-10", "10-20", "20-30", "30-40", "40-50", "50-60", "60-70", "70-80", "80-90", "90-100"], "f_flowedu_college": ["0-10", "10-20", "20-30", "30-40", "40-50", "50-60", "60-70", "70-80", "80-90", "90-100"], "f_flowedu_postgraduate": ["0-10", "10-20", "20-30", "30-40", "40-50", "50-60", "60-70", "70-80", "80-90", "90-100"], "f_flowedu_senior_middle": ["0-10", "10-20", "20-30", "30-40", "40-50", "50-60", "60-70", "70-80", "80-90", "90-100"], "f_flowhouse_middle": ["0-10", "10-20", "20-30", "30-40", "40-50", "50-60", "60-70", "70-80", "80-90", "90-100"], "f_flowhouse_low": ["0-10", "10-20", "20-30", "30-40", "40-50", "50-60", "60-70", "70-80", "80-90", "90-100"], "f_flowhouse_lower": ["0-10", "10-20", "20-30", "30-40", "40-50", "50-60", "60-70", "70-80", "80-90", "90-100"], "f_flowhouse_second_high": ["0-10", "10-20", "20-30", "30-40", "40-50", "50-60", "60-70", "70-80", "80-90", "90-100"], "f_flowhouse_high": ["0-10", "10-20", "20-30", "30-40", "40-50", "50-60", "60-70", "70-80", "80-90", "90-100"], "f_flowage_17": ["0-10", "10-20", "20-30", "30-40", "40-50", "50-60", "60-70", "70-80", "80-90", "90-100"], "f_flowage_24": ["0-10", "10-20", "20-30", "30-40", "40-50", "50-60", "60-70", "70-80", "80-90", "90-100"], "f_flowage_30": ["0-10", "10-20", "20-30", "30-40", "40-50", "50-60", "60-70", "70-80", "80-90", "90-100"], "f_flowage_35": ["0-10", "10-20", "20-30", "30-40", "40-50", "50-60", "60-70", "70-80", "80-90", "90-100"], "f_flowage_40": ["0-10", "10-20", "20-30", "30-40", "40-50", "50-60", "60-70", "70-80", "80-90", "90-100"], "f_flowage_45": ["0-10", "10-20", "20-30", "30-40", "40-50", "50-60", "60-70", "70-80", "80-90", "90-100"], "f_flowage_60": ["0-10", "10-20", "20-30", "30-40", "40-50", "50-60", "60-70", "70-80", "80-90", "90-100"], "f_flowage_over_60": ["0-10", "10-20", "20-30", "30-40", "40-50", "50-60", "60-70", "70-80", "80-90", "90-100"], "f_flowsex_woman": ["0-10", "10-20", "20-30", "30-40", "40-50", "50-60", "60-70", "70-80", "80-90", "90-100"], "f_flowsex_man": ["0-10", "10-20", "20-30", "30-40", "40-50", "50-60", "60-70", "70-80", "80-90", "90-100"], "f_holidayurban_cons_middle": ["0-10", "10-20", "20-30", "30-40", "40-50", "50-60", "60-70", "70-80", "80-90", "90-100"], "f_holidayurban_cons_low": ["0-10", "10-20", "20-30", "30-40", "40-50", "50-60", "60-70", "70-80", "80-90", "90-100"], "f_holidayurban_cons_lower": ["0-10", "10-20", "20-30", "30-40", "40-50", "50-60", "60-70", "70-80", "80-90", "90-100"], "f_holidayurban_cons_secondhigh": ["0-10", "10-20", "20-30", "30-40", "40-50", "50-60", "60-70", "70-80", "80-90", "90-100"], "f_holidayurban_cons_high": ["0-10", "10-20", "20-30", "30-40", "40-50", "50-60", "60-70", "70-80", "80-90", "90-100"], "f_holidayedu_junior_middle": ["0-10", "10-20", "20-30", "30-40", "40-50", "50-60", "60-70", "70-80", "80-90", "90-100"], "f_holidayedu_doctor": ["0-10", "10-20", "20-30", "30-40", "40-50", "50-60", "60-70", "70-80", "80-90", "90-100"], "f_holidayedu_specialty": ["0-10", "10-20", "20-30", "30-40", "40-50", "50-60", "60-70", "70-80", "80-90", "90-100"], "f_holidayedu_primary": ["0-10", "10-20", "20-30", "30-40", "40-50", "50-60", "60-70", "70-80", "80-90", "90-100"], "f_holidayedu_college": ["0-10", "10-20", "20-30", "30-40", "40-50", "50-60", "60-70", "70-80", "80-90", "90-100"], "f_holidayedu_postgraduate": ["0-10", "10-20", "20-30", "30-40", "40-50", "50-60", "60-70", "70-80", "80-90", "90-100"], "f_holidayedu_senior_middle": ["0-10", "10-20", "20-30", "30-40", "40-50", "50-60", "60-70", "70-80", "80-90", "90-100"], "f_holidayhouse_middle": ["0-10", "10-20", "20-30", "30-40", "40-50", "50-60", "60-70", "70-80", "80-90", "90-100"], "f_holidayhouse_low": ["0-10", "10-20", "20-30", "30-40", "40-50", "50-60", "60-70", "70-80", "80-90", "90-100"], "f_holidayhouse_lower": ["0-10", "10-20", "20-30", "30-40", "40-50", "50-60", "60-70", "70-80", "80-90", "90-100"], "f_holidayhouse_second_high": ["0-10", "10-20", "20-30", "30-40", "40-50", "50-60", "60-70", "70-80", "80-90", "90-100"], "f_holidayhouse_high": ["0-10", "10-20", "20-30", "30-40", "40-50", "50-60", "60-70", "70-80", "80-90", "90-100"], "f_holidayage_17": ["0-10", "10-20", "20-30", "30-40", "40-50", "50-60", "60-70", "70-80", "80-90", "90-100"], "f_holidayage_24": ["0-10", "10-20", "20-30", "30-40", "40-50", "50-60", "60-70", "70-80", "80-90", "90-100"], "f_holidayage_30": ["0-10", "10-20", "20-30", "30-40", "40-50", "50-60", "60-70", "70-80", "80-90", "90-100"], "f_holidayage_35": ["0-10", "10-20", "20-30", "30-40", "40-50", "50-60", "60-70", "70-80", "80-90", "90-100"], "f_holidayage_40": ["0-10", "10-20", "20-30", "30-40", "40-50", "50-60", "60-70", "70-80", "80-90", "90-100"], "f_holidayage_45": ["0-10", "10-20", "20-30", "30-40", "40-50", "50-60", "60-70", "70-80", "80-90", "90-100"], "f_holidayage_60": ["0-10", "10-20", "20-30", "30-40", "40-50", "50-60", "60-70", "70-80", "80-90", "90-100"], "f_holidayage_over_60": ["0-10", "10-20", "20-30", "30-40", "40-50", "50-60", "60-70", "70-80", "80-90", "90-100"], "f_holidaysex_woman": ["0-10", "10-20", "20-30", "30-40", "40-50", "50-60", "60-70", "70-80", "80-90", "90-100"], "f_holidaysex_man": ["0-10", "10-20", "20-30", "30-40", "40-50", "50-60", "60-70", "70-80", "80-90", "90-100"], "f_workday_urban_cons_middle": ["0-10", "10-20", "20-30", "30-40", "40-50", "50-60", "60-70", "70-80", "80-90", "90-100"], "f_workday_urban_cons_low": ["0-10", "10-20", "20-30", "30-40", "40-50", "50-60", "60-70", "70-80", "80-90", "90-100"], "f_workday_urban_cons_lower": ["0-10", "10-20", "20-30", "30-40", "40-50", "50-60", "60-70", "70-80", "80-90", "90-100"], "f_workday_urban_cons_secondhigh": ["0-10", "10-20", "20-30", "30-40", "40-50", "50-60", "60-70", "70-80", "80-90", "90-100"], "f_workday_urban_cons_high": ["0-10", "10-20", "20-30", "30-40", "40-50", "50-60", "60-70", "70-80", "80-90", "90-100"], "f_workday_edu_junior_middle": ["0-10", "10-20", "20-30", "30-40", "40-50", "50-60", "60-70", "70-80", "80-90", "90-100"], "f_workday_edu_doctor": ["0-10", "10-20", "20-30", "30-40", "40-50", "50-60", "60-70", "70-80", "80-90", "90-100"], "f_workday_edu_specialty": ["0-10", "10-20", "20-30", "30-40", "40-50", "50-60", "60-70", "70-80", "80-90", "90-100"], "f_workday_edu_primary": ["0-10", "10-20", "20-30", "30-40", "40-50", "50-60", "60-70", "70-80", "80-90", "90-100"], "f_workday_edu_college": ["0-10", "10-20", "20-30", "30-40", "40-50", "50-60", "60-70", "70-80", "80-90", "90-100"], "f_workday_edu_postgraduate": ["0-10", "10-20", "20-30", "30-40", "40-50", "50-60", "60-70", "70-80", "80-90", "90-100"], "f_workday_edu_senior_middle": ["0-10", "10-20", "20-30", "30-40", "40-50", "50-60", "60-70", "70-80", "80-90", "90-100"], "f_workday_house_middle": ["0-10", "10-20", "20-30", "30-40", "40-50", "50-60", "60-70", "70-80", "80-90", "90-100"], "f_workday_house_low": ["0-10", "10-20", "20-30", "30-40", "40-50", "50-60", "60-70", "70-80", "80-90", "90-100"], "f_workday_house_lower": ["0-10", "10-20", "20-30", "30-40", "40-50", "50-60", "60-70", "70-80", "80-90", "90-100"], "f_workday_house_second_high": ["0-10", "10-20", "20-30", "30-40", "40-50", "50-60", "60-70", "70-80", "80-90", "90-100"], "f_workday_house_high": ["0-10", "10-20", "20-30", "30-40", "40-50", "50-60", "60-70", "70-80", "80-90", "90-100"], "f_workday_age_17": ["0-10", "10-20", "20-30", "30-40", "40-50", "50-60", "60-70", "70-80", "80-90", "90-100"], "f_workday_age_24": ["0-10", "10-20", "20-30", "30-40", "40-50", "50-60", "60-70", "70-80", "80-90", "90-100"], "f_workday_age_30": ["0-10", "10-20", "20-30", "30-40", "40-50", "50-60", "60-70", "70-80", "80-90", "90-100"], "f_workday_age_35": ["0-10", "10-20", "20-30", "30-40", "40-50", "50-60", "60-70", "70-80", "80-90", "90-100"], "f_workday_age_40": ["0-10", "10-20", "20-30", "30-40", "40-50", "50-60", "60-70", "70-80", "80-90", "90-100"], "f_workday_age_45": ["0-10", "10-20", "20-30", "30-40", "40-50", "50-60", "60-70", "70-80", "80-90", "90-100"], "f_workday_age_60": ["0-10", "10-20", "20-30", "30-40", "40-50", "50-60", "60-70", "70-80", "80-90", "90-100"], "f_workday_age_over_60": ["0-10", "10-20", "20-30", "30-40", "40-50", "50-60", "60-70", "70-80", "80-90", "90-100"], "f_workday_sex_woman": ["0-10", "10-20", "20-30", "30-40", "40-50", "50-60", "60-70", "70-80", "80-90", "90-100"], "f_workday_sex_man": ["0-10", "10-20", "20-30", "30-40", "40-50", "50-60", "60-70", "70-80", "80-90", "90-100"], } CLEANING_RULES = { "r_home_num": {"method": "fillna", "opt": "fill", "value": "501-2000", "type": "str"}, "r_work_num": {"method": "fillna", "opt": "fill", "value": "501-2000", "type": "str"}, "r_resident_num": {"method": "fillna", "opt": "fill", "value": "501-2000", "type": "str"}, "r_urban_cons_middle": {"method": "fillna", "opt": "fill", "value": "40-50", "type": "str"}, "r_urban_cons_low": {"method": "fillna", "opt": "fill", "value": "40-50", "type": "str"}, "r_urban_cons_lower": {"method": "fillna", "opt": "fill", "value": "40-50", "type": "str"}, "r_urban_cons_secondhigh": {"method": "fillna", "opt": "fill", "value": "40-50", "type": "str"}, "r_urban_cons_high": {"method": "fillna", "opt": "fill", "value": "40-50", "type": "str"}, "r_edu_junior_middle": {"method": "fillna", "opt": "fill", "value": "40-50", "type": "str"}, "r_edu_doctor": {"method": "fillna", "opt": "fill", "value": "40-50", "type": "str"}, "r_edu_specialty": {"method": "fillna", "opt": "fill", "value": "40-50", "type": "str"}, "r_edu_primary": {"method": "fillna", "opt": "fill", "value": "40-50", "type": "str"}, "r_edu_college": {"method": "fillna", "opt": "fill", "value": "40-50", "type": "str"}, "r_edu_postgraduate": {"method": "fillna", "opt": "fill", "value": "40-50", "type": "str"}, "r_edu_senior_middle": {"method": "fillna", "opt": "fill", "value": "40-50", "type": "str"}, "r_house_price79999": {"method": "fillna", "opt": "fill", "value": "40-50", "type": "str"}, "r_house_price59999": {"method": "fillna", "opt": "fill", "value": "40-50", "type": "str"}, "r_house_price39999": {"method": "fillna", "opt": "fill", "value": "40-50", "type": "str"}, "r_house_price19999": {"method": "fillna", "opt": "fill", "value": "40-50", "type": "str"}, "r_house_price9999": {"method": "fillna", "opt": "fill", "value": "40-50", "type": "str"}, "r_house_price7999": {"method": "fillna", "opt": "fill", "value": "40-50", "type": "str"}, "r_house_price4999": {"method": "fillna", "opt": "fill", "value": "40-50", "type": "str"}, "r_age_17": {"method": "fillna", "opt": "fill", "value": "40-50", "type": "str"}, "r_age_24": {"method": "fillna", "opt": "fill", "value": "40-50", "type": "str"}, "r_age_30": {"method": "fillna", "opt": "fill", "value": "40-50", "type": "str"}, "r_age_35": {"method": "fillna", "opt": "fill", "value": "40-50", "type": "str"}, "r_age_40": {"method": "fillna", "opt": "fill", "value": "40-50", "type": "str"}, "r_age_45": {"method": "fillna", "opt": "fill", "value": "40-50", "type": "str"}, "r_age_60": {"method": "fillna", "opt": "fill", "value": "40-50", "type": "str"}, "r_age_over_60": {"method": "fillna", "opt": "fill", "value": "40-50", "type": "str"}, "r_sex_woman": {"method": "fillna", "opt": "fill", "value": "40-50", "type": "str"}, "r_sex_man": {"method": "fillna", "opt": "fill", "value": "40-50", "type": "str"}, "r_catering_50": {"method": "fillna", "opt": "fill", "value": "40-50", "type": "str"}, "r_catering_100": {"method": "fillna", "opt": "fill", "value": "40-50", "type": "str"}, "r_catering_150": {"method": "fillna", "opt": "fill", "value": "40-50", "type": "str"}, "r_catering_200": {"method": "fillna", "opt": "fill", "value": "40-50", "type": "str"}, "r_catering_500": {"method": "fillna", "opt": "fill", "value": "40-50", "type": "str"}, "r_catering_over_500": {"method": "fillna", "opt": "fill", "value": "40-50", "type": "str"}, "r_catering_times_2": {"method": "fillna", "opt": "fill", "value": "40-50", "type": "str"}, "r_catering_times_4": {"method": "fillna", "opt": "fill", "value": "40-50", "type": "str"}, "r_catering_times_6": {"method": "fillna", "opt": "fill", "value": "40-50", "type": "str"}, "r_catering_times_8": {"method": "fillna", "opt": "fill", "value": "40-50", "type": "str"}, "r_catering_times_10": {"method": "fillna", "opt": "fill", "value": "40-50", "type": "str"}, "r_catering_times_11": {"method": "fillna", "opt": "fill", "value": "40-50", "type": "str"}, "r_native_beijing": {"method": "fillna", "opt": "fill", "value": "40-50", "type": "str"}, "r_native_tianjing": {"method": "fillna", "opt": "fill", "value": "40-50", "type": "str"}, "r_native_hebei": {"method": "fillna", "opt": "fill", "value": "40-50", "type": "str"}, "r_native_shanxi": {"method": "fillna", "opt": "fill", "value": "40-50", "type": "str"}, "r_native_neimeng": {"method": "fillna", "opt": "fill", "value": "40-50", "type": "str"}, "r_native_liaoning": {"method": "fillna", "opt": "fill", "value": "40-50", "type": "str"}, "r_native_jilin": {"method": "fillna", "opt": "fill", "value": "40-50", "type": "str"}, "r_native_heilongjiang": {"method": "fillna", "opt": "fill", "value": "40-50", "type": "str"}, "r_native_shanghai": {"method": "fillna", "opt": "fill", "value": "40-50", "type": "str"}, "r_native_jiangsu": {"method": "fillna", "opt": "fill", "value": "40-50", "type": "str"}, "r_native_zhejiang": {"method": "fillna", "opt": "fill", "value": "40-50", "type": "str"}, "r_native_anhui": {"method": "fillna", "opt": "fill", "value": "40-50", "type": "str"}, "r_native_fujian": {"method": "fillna", "opt": "fill", "value": "40-50", "type": "str"}, "r_native_jiangix": {"method": "fillna", "opt": "fill", "value": "40-50", "type": "str"}, "r_native_shandong": {"method": "fillna", "opt": "fill", "value": "40-50", "type": "str"}, "r_native_henan": {"method": "fillna", "opt": "fill", "value": "40-50", "type": "str"}, "r_native_hubei": {"method": "fillna", "opt": "fill", "value": "40-50", "type": "str"}, "r_native_hunan": {"method": "fillna", "opt": "fill", "value": "40-50", "type": "str"}, "r_native_guangdong": {"method": "fillna", "opt": "fill", "value": "40-50", "type": "str"}, "r_native_hainan": {"method": "fillna", "opt": "fill", "value": "40-50", "type": "str"}, "r_native_sichuan": {"method": "fillna", "opt": "fill", "value": "40-50", "type": "str"}, "r_native_guizhou": {"method": "fillna", "opt": "fill", "value": "40-50", "type": "str"}, "r_native_yunnan": {"method": "fillna", "opt": "fill", "value": "40-50", "type": "str"}, "r_native_shan": {"method": "fillna", "opt": "fill", "value": "40-50", "type": "str"}, "r_native_gansu": {"method": "fillna", "opt": "fill", "value": "40-50", "type": "str"}, "r_native_qinghai": {"method": "fillna", "opt": "fill", "value": "40-50", "type": "str"}, "r_native_guangxi": {"method": "fillna", "opt": "fill", "value": "40-50", "type": "str"}, "r_native_ningxia": {"method": "fillna", "opt": "fill", "value": "40-50", "type": "str"}, "r_native_xinjiang": {"method": "fillna", "opt": "fill", "value": "40-50", "type": "str"}, "r_native_xizang": {"method": "fillna", "opt": "fill", "value": "40-50", "type": "str"}, "r_native_chongqing": {"method": "fillna", "opt": "fill", "value": "40-50", "type": "str"}, "r_native_hongkong": {"method": "fillna", "opt": "fill", "value": "40-50", "type": "str"}, "r_native_macao": {"method": "fillna", "opt": "fill", "value": "40-50", "type": "str"}, "r_native_taiwan": {"method": "fillna", "opt": "fill", "value": "40-50", "type": "str"}, "r_native_other": {"method": "fillna", "opt": "fill", "value": "40-50", "type": "str"}, "f_flow_num": {"method": "fillna", "opt": "fill", "value": "2001-5000", "type": "str"}, "f_holiday_flow_num": {"method": "fillna", "opt": "fill", "value": "2001-5000", "type": "str"}, "f_workday_flow_num": {"method": "fillna", "opt": "fill", "value": "2001-5000", "type": "str"}, "f_flowurban_cons_middle": {"method": "fillna", "opt": "fill", "value": "40-50", "type": "str"}, "f_flowurban_cons_low": {"method": "fillna", "opt": "fill", "value": "40-50", "type": "str"}, "f_flowurban_cons_lower": {"method": "fillna", "opt": "fill", "value": "40-50", "type": "str"}, "f_flowurban_cons_second_high": {"method": "fillna", "opt": "fill", "value": "40-50", "type": "str"}, "f_flowurban_cons_high": {"method": "fillna", "opt": "fill", "value": "40-50", "type": "str"}, "f_flowedu_junior_middle": {"method": "fillna", "opt": "fill", "value": "40-50", "type": "str"}, "f_flowedu_doctor": {"method": "fillna", "opt": "fill", "value": "40-50", "type": "str"}, "f_flowedu_specialty": {"method": "fillna", "opt": "fill", "value": "40-50", "type": "str"}, "f_flowedu_primary": {"method": "fillna", "opt": "fill", "value": "40-50", "type": "str"}, "f_flowedu_college": {"method": "fillna", "opt": "fill", "value": "40-50", "type": "str"}, "f_flowedu_postgraduate": {"method": "fillna", "opt": "fill", "value": "40-50", "type": "str"}, "f_flowedu_senior_middle": {"method": "fillna", "opt": "fill", "value": "40-50", "type": "str"}, "f_flowhouse_middle": {"method": "fillna", "opt": "fill", "value": "40-50", "type": "str"}, "f_flowhouse_low": {"method": "fillna", "opt": "fill", "value": "40-50", "type": "str"}, "f_flowhouse_lower": {"method": "fillna", "opt": "fill", "value": "40-50", "type": "str"}, "f_flowhouse_second_high": {"method": "fillna", "opt": "fill", "value": "40-50", "type": "str"}, "f_flowhouse_high": {"method": "fillna", "opt": "fill", "value": "40-50", "type": "str"}, "f_flowage_17": {"method": "fillna", "opt": "fill", "value": "40-50", "type": "str"}, "f_flowage_24": {"method": "fillna", "opt": "fill", "value": "40-50", "type": "str"}, "f_flowage_30": {"method": "fillna", "opt": "fill", "value": "40-50", "type": "str"}, "f_flowage_35": {"method": "fillna", "opt": "fill", "value": "40-50", "type": "str"}, "f_flowage_40": {"method": "fillna", "opt": "fill", "value": "40-50", "type": "str"}, "f_flowage_45": {"method": "fillna", "opt": "fill", "value": "40-50", "type": "str"}, "f_flowage_60": {"method": "fillna", "opt": "fill", "value": "40-50", "type": "str"}, "f_flowage_over_60": {"method": "fillna", "opt": "fill", "value": "40-50", "type": "str"}, "f_flowsex_woman": {"method": "fillna", "opt": "fill", "value": "40-50", "type": "str"}, "f_flowsex_man": {"method": "fillna", "opt": "fill", "value": "40-50", "type": "str"}, "f_holidayurban_cons_middle": {"method": "fillna", "opt": "fill", "value": "40-50", "type": "str"}, "f_holidayurban_cons_low": {"method": "fillna", "opt": "fill", "value": "40-50", "type": "str"}, "f_holidayurban_cons_lower": {"method": "fillna", "opt": "fill", "value": "40-50", "type": "str"}, "f_holidayurban_cons_secondhigh": {"method": "fillna", "opt": "fill", "value": "40-50", "type": "str"}, "f_holidayurban_cons_high": {"method": "fillna", "opt": "fill", "value": "40-50", "type": "str"}, "f_holidayedu_junior_middle": {"method": "fillna", "opt": "fill", "value": "40-50", "type": "str"}, "f_holidayedu_doctor": {"method": "fillna", "opt": "fill", "value": "40-50", "type": "str"}, "f_holidayedu_specialty": {"method": "fillna", "opt": "fill", "value": "40-50", "type": "str"}, "f_holidayedu_primary": {"method": "fillna", "opt": "fill", "value": "40-50", "type": "str"}, "f_holidayedu_college": {"method": "fillna", "opt": "fill", "value": "40-50", "type": "str"}, "f_holidayedu_postgraduate": {"method": "fillna", "opt": "fill", "value": "40-50", "type": "str"}, "f_holidayedu_senior_middle": {"method": "fillna", "opt": "fill", "value": "40-50", "type": "str"}, "f_holidayhouse_middle": {"method": "fillna", "opt": "fill", "value": "40-50", "type": "str"}, "f_holidayhouse_low": {"method": "fillna", "opt": "fill", "value": "40-50", "type": "str"}, "f_holidayhouse_lower": {"method": "fillna", "opt": "fill", "value": "40-50", "type": "str"}, "f_holidayhouse_second_high": {"method": "fillna", "opt": "fill", "value": "40-50", "type": "str"}, "f_holidayhouse_high": {"method": "fillna", "opt": "fill", "value": "40-50", "type": "str"}, "f_holidayage_17": {"method": "fillna", "opt": "fill", "value": "40-50", "type": "str"}, "f_holidayage_24": {"method": "fillna", "opt": "fill", "value": "40-50", "type": "str"}, "f_holidayage_30": {"method": "fillna", "opt": "fill", "value": "40-50", "type": "str"}, "f_holidayage_35": {"method": "fillna", "opt": "fill", "value": "40-50", "type": "str"}, "f_holidayage_40": {"method": "fillna", "opt": "fill", "value": "40-50", "type": "str"}, "f_holidayage_45": {"method": "fillna", "opt": "fill", "value": "40-50", "type": "str"}, "f_holidayage_60": {"method": "fillna", "opt": "fill", "value": "40-50", "type": "str"}, "f_holidayage_over_60": {"method": "fillna", "opt": "fill", "value": "40-50", "type": "str"}, "f_holidaysex_woman": {"method": "fillna", "opt": "fill", "value": "40-50", "type": "str"}, "f_holidaysex_man": {"method": "fillna", "opt": "fill", "value": "40-50", "type": "str"}, "f_workday_urban_cons_middle": {"method": "fillna", "opt": "fill", "value": "40-50", "type": "str"}, "f_workday_urban_cons_low": {"method": "fillna", "opt": "fill", "value": "40-50", "type": "str"}, "f_workday_urban_cons_lower": {"method": "fillna", "opt": "fill", "value": "40-50", "type": "str"}, "f_workday_urban_cons_secondhigh": {"method": "fillna", "opt": "fill", "value": "40-50", "type": "str"}, "f_workday_urban_cons_high": {"method": "fillna", "opt": "fill", "value": "40-50", "type": "str"}, "f_workday_edu_junior_middle": {"method": "fillna", "opt": "fill", "value": "40-50", "type": "str"}, "f_workday_edu_doctor": {"method": "fillna", "opt": "fill", "value": "40-50", "type": "str"}, "f_workday_edu_specialty": {"method": "fillna", "opt": "fill", "value": "40-50", "type": "str"}, "f_workday_edu_primary": {"method": "fillna", "opt": "fill", "value": "40-50", "type": "str"}, "f_workday_edu_college": {"method": "fillna", "opt": "fill", "value": "40-50", "type": "str"}, "f_workday_edu_postgraduate": {"method": "fillna", "opt": "fill", "value": "40-50", "type": "str"}, "f_workday_edu_senior_middle": {"method": "fillna", "opt": "fill", "value": "40-50", "type": "str"}, "f_workday_house_middle": {"method": "fillna", "opt": "fill", "value": "40-50", "type": "str"}, "f_workday_house_low": {"method": "fillna", "opt": "fill", "value": "40-50", "type": "str"}, "f_workday_house_lower": {"method": "fillna", "opt": "fill", "value": "40-50", "type": "str"}, "f_workday_house_second_high": {"method": "fillna", "opt": "fill", "value": "40-50", "type": "str"}, "f_workday_house_high": {"method": "fillna", "opt": "fill", "value": "40-50", "type": "str"}, "f_workday_age_17": {"method": "fillna", "opt": "fill", "value": "40-50", "type": "str"}, "f_workday_age_24": {"method": "fillna", "opt": "fill", "value": "40-50", "type": "str"}, "f_workday_age_30": {"method": "fillna", "opt": "fill", "value": "40-50", "type": "str"}, "f_workday_age_35": {"method": "fillna", "opt": "fill", "value": "40-50", "type": "str"}, "f_workday_age_40": {"method": "fillna", "opt": "fill", "value": "40-50", "type": "str"}, "f_workday_age_45": {"method": "fillna", "opt": "fill", "value": "40-50", "type": "str"}, "f_workday_age_60": {"method": "fillna", "opt": "fill", "value": "40-50", "type": "str"}, "f_workday_age_over_60": {"method": "fillna", "opt": "fill", "value": "40-50", "type": "str"}, "f_workday_sex_woman": {"method": "fillna", "opt": "fill", "value": "40-50", "type": "str"}, "f_workday_sex_man": {"method": "fillna", "opt": "fill", "value": "40-50", "type": "str"}, } class ImportanceFeaturesMap: CUSTOM_FEATURES_MAP = { "BB_RTL_CUST_MARKET_TYPE_NAME": "零售户市场类型名称", "BB_RTL_CUST_BUSINESS_TYPE_NAME": "零售客户业态名称", "BB_RTL_CUST_CHAIN_FLAG": "零售户连锁标识", "MD04_MG_RTL_CUST_CREDITCLASS_NAME": "零售户信用等级名称", "MD04_DIR_SAL_STORE_FLAG": "直营店标识", "BB_CUSTOMER_MANAGER_SCOPE_NAME": "零售户经营范围名称", "BB_RTL_CUST_TERMINAL_LEVEL_NAME": "零售户终端层级名称", "OPERATOR_EDU": "零售客户经营者文化程度", "STORE_AREA": "店铺经营面积", "OPERATOR_AGE": "经营者年龄", "PRODUCT_INSALE_QTY": "在销品规数", } PRODUCT_FEATRUES_MAP = { "factory_name": "产地", "brand_name": "品牌名称", "is_low_tar": "低焦油卷烟", "is_medium": "中支烟", "is_tiny": "细支烟", "is_coarse": "粗支烟", "is_exploding_beads": "爆珠烟", "is_abnormity": "异形包装", "is_cig": "雪茄烟", "is_chuangxin": "创新品类", "direct_retail_price": "卷烟建议零售价", "tbc_total_length": "烟支总长度", "product_style": "包装类型", } SHOPING_FEATURES_MAP = { "r_home_num": "常驻人口_居住人数", "r_work_num": "常驻人口_工作人数", "r_resident_num": "常驻人口_工作或居住人数", "r_urban_cons_middle": "常驻人口_城市消费水平_中", "r_urban_cons_low": "常驻人口_城市消费水平_低", "r_urban_cons_lower": "常驻人口_城市消费水平_次低", "r_urban_cons_secondhigh": "常驻人口_城市消费水平_次高", "r_urban_cons_high": "常驻人口_城市消费水平_高", "r_edu_junior_middle": "常驻人口_学历_初中", "r_edu_doctor": "常驻人口_学历_博士", "r_edu_specialty": "常驻人口_学历_大专", "r_edu_primary": "常驻人口_学历_小学", "r_edu_college": "常驻人口_学历_本科", "r_edu_postgraduate": "常驻人口_学历_硕士", "r_edu_senior_middle": "常驻人口_学历_高中", "r_house_price79999": "常驻人口_居住社区房价_60000_79999", "r_house_price59999": "常驻人口_居住社区房价_40000_59999", "r_house_price39999": "常驻人口_居住社区房价_20000_39999", "r_house_price19999": "常驻人口_居住社区房价_10000_19999", "r_house_price9999": "常驻人口_居住社区房价_8000_9999", "r_house_price7999": "常驻人口_居住社区房价_5000_7999", "r_house_price4999": "常驻人口_居住社区房价_2000_4999", "r_age_17": "常驻人口_年龄_0_17", "r_age_24": "常驻人口_年龄_18_24", "r_age_30": "常驻人口_年龄_25_30", "r_age_35": "常驻人口_年龄_31_35", "r_age_40": "常驻人口_年龄_36_40", "r_age_45": "常驻人口_年龄_41_45", "r_age_60": "常驻人口_年龄_46_60", "r_age_over_60": "常驻人口_年龄_61以上", "r_sex_woman": "常驻人口_性别_女", "r_sex_man": "常驻人口_性别_男", "r_catering_50": "常驻人口_餐饮消费水平_50", "r_catering_100": "常驻人口_餐饮消费水平_100", "r_catering_150": "常驻人口_餐饮消费水平_150", "r_catering_200": "常驻人口_餐饮消费水平_200", "r_catering_500": "常驻人口_餐饮消费水平_500", "r_catering_over_500": "常驻人口_餐饮消费水平_500以上", "r_catering_times_2": "常驻人口_餐饮消费频次_1_2", "r_catering_times_4": "常驻人口_餐饮消费频次_2_4", "r_catering_times_6": "常驻人口_餐饮消费频次_4_6", "r_catering_times_8": "常驻人口_餐饮消费频次_6_8", "r_catering_times_10": "常驻人口_餐饮消费频次_8_10", "r_catering_times_11": "常驻人口_餐饮消费频次_11以上", "r_native_beijing": "常驻人口_家乡地_北京市", "r_native_tianjing": "常驻人口_家乡地_天津市", "r_native_hebei": "常驻人口_家乡地_河北省", "r_native_shanxi": "常驻人口_家乡地_山西省", "r_native_neimeng": "常驻人口_家乡地_内蒙古", "r_native_liaoning": "常驻人口_家乡地_辽宁省", "r_native_jilin": "常驻人口_家乡地_吉林省", "r_native_heilongjiang": "常驻人口_家乡地_黑龙江省", "r_native_shanghai": "常驻人口_家乡地_上海市", "r_native_jiangsu": "常驻人口_家乡地_江苏省", "r_native_zhejiang": "常驻人口_家乡地_浙江省", "r_native_anhui": "常驻人口_家乡地_安徽省", "r_native_fujian": "常驻人口_家乡地_福建省", "r_native_jiangix": "常驻人口_家乡地_江西省", "r_native_shandong": "常驻人口_家乡地_山东省", "r_native_henan": "常驻人口_家乡地_河南省", "r_native_hubei": "常驻人口_家乡地_湖北省", "r_native_hunan": "常驻人口_家乡地_湖南省", "r_native_guangdong": "常驻人口_家乡地_广东省", "r_native_hainan": "常驻人口_家乡地_海南省", "r_native_sichuan": "常驻人口_家乡地_四川省", "r_native_guizhou": "常驻人口_家乡地_贵州省", "r_native_yunnan": "常驻人口_家乡地_云南省", "r_native_shan": "常驻人口_家乡地_陕西省", "r_native_gansu": "常驻人口_家乡地_甘肃省", "r_native_qinghai": "常驻人口_家乡地_青海省", "r_native_guangxi": "常驻人口_家乡地_广西壮族自治区", "r_native_ningxia": "常驻人口_家乡地_宁夏回族自治区", "r_native_xinjiang": "常驻人口_家乡地_新疆维吾尔自治区", "r_native_xizang": "常驻人口_家乡地_西藏自治区", "r_native_chongqing": "常驻人口_家乡地_重庆市", "r_native_hongkong": "常驻人口_家乡地_香港", "r_native_macao": "常驻人口_家乡地_澳门", "r_native_taiwan": "常驻人口_家乡地_台湾", "r_native_other": "常驻人口_家乡地_其它", "f_flow_num": "流动人口_工作日_日均流动人口数量", "f_holiday_flow_num": "流动人口_节假日_日均流动人口数量", "f_workday_flow_num": "流动人口_日均流动人口数量", "f_flowurban_cons_middle": "日均流动_城市消费水平_中", "f_flowurban_cons_low": "日均流动_城市消费水平_低", "f_flowurban_cons_lower": "日均流动_城市消费水平_次低", "f_flowurban_cons_second_high": "日均流动_城市消费水平_次高", "f_flowurban_cons_high": "日均流动_城市消费水平_高", "f_flowedu_junior_middle": "日均流动_学历_初中", "f_flowedu_doctor": "日均流动_学历_博士", "f_flowedu_specialty": "日均流动_学历_大专", "f_flowedu_primary": "日均流动_学历_小学", "f_flowedu_college": "日均流动_学历_本科", "f_flowedu_postgraduate": "日均流动_学历_硕士", "f_flowedu_senior_middle": "日均流动_学历_高中", "f_flowhouse_middle": "日均流动_居住社区房价_中", "f_flowhouse_low": "日均流动_居住社区房价_低", "f_flowhouse_lower": "日均流动_居住社区房价_次低", "f_flowhouse_second_high": "日均流动_居住社区房价_次高", "f_flowhouse_high": "日均流动_居住社区房价_高", "f_flowage_17": "日均流动_年龄_0_17", "f_flowage_24": "日均流动_年龄_18_24", "f_flowage_30": "日均流动_年龄_25_30", "f_flowage_35": "日均流动_年龄_31_35", "f_flowage_40": "日均流动_年龄_36_40", "f_flowage_45": "日均流动_年龄_41_45", "f_flowage_60": "日均流动_年龄_46_60", "f_flowage_over_60": "日均流动_年龄_61以上", "f_flowsex_woman": "日均流动_性别_女", "f_flowsex_man": "日均流动_性别_男", "f_holidayurban_cons_middle": "节假日流动_城市消费水平_中", "f_holidayurban_cons_low": "节假日流动_城市消费水平_低", "f_holidayurban_cons_lower": "节假日流动_城市消费水平_次低", "f_holidayurban_cons_secondhigh": "节假日流动_城市消费水平_次高", "f_holidayurban_cons_high": "节假日流动_城市消费水平_高", "f_holidayedu_junior_middle": "节假日流动_学历_初中", "f_holidayedu_doctor": "节假日流动_学历_博士", "f_holidayedu_specialty": "节假日流动_学历_大专", "f_holidayedu_primary": "节假日流动_学历_小学", "f_holidayedu_college": "节假日流动_学历_本科", "f_holidayedu_postgraduate": "节假日流动_学历_硕士", "f_holidayedu_senior_middle": "节假日流动_学历_高中", "f_holidayhouse_middle": "节假日流动_居住社区房价_中", "f_holidayhouse_low": "节假日流动_居住社区房价_低", "f_holidayhouse_lower": "节假日流动_居住社区房价_次低", "f_holidayhouse_second_high": "节假日流动_居住社区房价_次高", "f_holidayhouse_high": "节假日流动_居住社区房价_高", "f_holidayage_17": "节假日流动_年龄_0_17", "f_holidayage_24": "节假日流动_年龄_18_24", "f_holidayage_30": "节假日流动_年龄_25_30", "f_holidayage_35": "节假日流动_年龄_31_35", "f_holidayage_40": "节假日流动_年龄_36_40", "f_holidayage_45": "节假日流动_年龄_41_45", "f_holidayage_60": "节假日流动_年龄_46_60", "f_holidayage_over_60": "节假日流动_年龄_61以上", "f_holidaysex_woman": "节假日流动_性别_女", "f_holidaysex_man": "节假日流动_性别_男", "f_workday_urban_cons_middle": "工作日流动_城市消费水平_中", "f_workday_urban_cons_low": "工作日流动_城市消费水平_低", "f_workday_urban_cons_lower": "工作日流动_城市消费水平_次低", "f_workday_urban_cons_secondhigh": "工作日流动_城市消费水平_次高", "f_workday_urban_cons_high": "工作日流动_城市消费水平_高", "f_workday_edu_junior_middle": "工作日流动_学历_初中", "f_workday_edu_doctor": "工作日流动_学历_博士", "f_workday_edu_specialty": "工作日流动_学历_大专", "f_workday_edu_primary": "工作日流动_学历_小学", "f_workday_edu_college": "工作日流动_学历_本科", "f_workday_edu_postgraduate": "工作日流动_学历_硕士", "f_workday_edu_senior_middle": "工作日流动_学历_高中", "f_workday_house_middle": "工作日流动_居住社区房价_中", "f_workday_house_low": "工作日流动_居住社区房价_低", "f_workday_house_lower": "工作日流动_居住社区房价_次低", "f_workday_house_second_high": "工作日流动_居住社区房价_次高", "f_workday_house_high": "工作日流动_居住社区房价_高", "f_workday_age_17": "工作日流动_年龄_0_17", "f_workday_age_24": "工作日流动_年龄_18_24", "f_workday_age_30": "工作日流动_年龄_25_30", "f_workday_age_35": "工作日流动_年龄_31_35", "f_workday_age_40": "工作日流动_年龄_36_40", "f_workday_age_45": "工作日流动_年龄_41_45", "f_workday_age_60": "工作日流动_年龄_46_60", "f_workday_age_over_60": "工作日流动_年龄_61以上", "f_workday_sex_woman": "工作日流动_性别_女", "f_workday_sex_man": "工作日流动_性别_男", } class DeliveryConfig: FEATURE_COLUMNS = [ "customer_code", # 零售户代码 "goods_code", # 卷烟代码 "retail_index_week", # 周市场零售价格监测指数 "turnover_rate_collpoint", # 采集点销售量动销率（周） "turnover_rate_terminal", # 零售终端销售量动销率（周） "sale_qty", # 周销售量 ] CLEANING_RULES = { "retail_index_week": {"method": "fillna", "opt": "fill", "value": 0.0000, "type": "num"}, "turnover_rate_collpoint": {"method": "fillna", "opt": "fill", "value": 0.0000, "type": "num"}, "turnover_rate_terminal": {"method": "fillna", "opt": "fill", "value": 0.0000, "type": "num"}, "sale_qty": {"method": "fillna", "opt": "fill", "value": 0, "type": "num"}, } FEATURES_MAP = { "retail_index_week": "市场零售价格监测指数（周）", "turnover_rate_collpoint": "采集点销售量动销率（周）", "turnover_rate_terminal": "零售终端销售量动销率（周）", "sale_qty": "周销售量" }