TIANCHI安泰杯 —跨境电商智能算法大赛Baseline

import numpy as np
import pandas as pd
import os 
from tqdm import tqdm_notebook
import lightgbm as lgb
import xgboost as xgb
import warnings
warnings.simplefilter(action='ignore', category=FutureWarning)
warnings.filterwarnings('ignore')

path = './input/'
item   = pd.read_csv(path+'Antai_AE_round1_item_attr_20190626.csv')
submit = pd.read_csv(path+'Antai_AE_round1_submit_20190715.csv', header=None)
test   = pd.read_csv(path+'Antai_AE_round1_test_20190626.csv')
train  = pd.read_csv(path+'Antai_AE_round1_train_20190626.csv')

def get_preprocessing(df_):
    df = df_.copy()   
    df['hour']  = df['create_order_time'].apply(lambda x:int(x[11:13]))
    df['day']   = df['create_order_time'].apply(lambda x:int(x[8:10]))
    df['month'] = df['create_order_time'].apply(lambda x:int(x[5:7]))
    df['year']  = df['create_order_time'].apply(lambda x:int(x[0:4]))
    df['date']  = (df['month'].values - 7) * 31 + df['day']    
    del df['create_order_time']    
    return df

train = get_preprocessing(train)
test  = get_preprocessing(test)

# 高频item_id
temp = train.loc[train.buyer_country_id=='yy']
temp = temp.drop_duplicates(subset=['buyer_admin_id','item_id'], keep='first')
item_cnts = temp.groupby(['item_id']).size().reset_index()
item_cnts.columns = ['item_id','cnts']
item_cnts = item_cnts.sort_values('cnts', ascending=False)
items = item_cnts['item_id'].values.tolist()

# 很多admin的历史行为不够30个item，所以就需要填充够30个
# 这里使用train下yy的数据构造item_id频次排序，然后依次填充
def item_fillna(tmp_):  
    tmp = tmp_.copy()   
    l = len(tmp)
    if l == 30:
        tmp = tmp
    elif l < 30:
        m = 30 - l
        items_t = items.copy()
        for i in range(m):
            for j in range(50):
                it = items_t.pop(0)
                if it not in tmp:
                    tmp.append(it)
                    break
    elif l > 30:
        tmp = tmp[:30]
    
    return tmp

# 获取top30的item
def get_item_list(df_):
    df = df_.copy()
    dic = {}
    flag = 0
    for item in df[['buyer_admin_id','item_id']].values:
        try:
            dic[item[0]].append(item[1])
        except:
            if flag != 0:
                # 去重
                tmp = []
                for i in dic[flag]:
                    if i not in tmp:
                        tmp.append(i)
                # 填充
                tmp = item_fillna(tmp)
                dic[flag] = tmp
                
                flag = item[0]
            else:
                flag = item[0]
            dic[item[0]] = [item[1]]

    return dic

test = test.sort_values(['buyer_admin_id','irank'])
dic = get_item_list(test)

# 最终提交
temp = pd.DataFrame({'lst':dic}).reset_index()
for i in range(30):
    temp[i] = temp['lst'].apply(lambda x:x[i])
del temp['lst']
temp.to_csv('submission.csv',index=False,header=None)