EduCoder_Study_RS/ranking/shixun/xdeepfm_ranker_predict.py

import os
import sys
sys.path.append(os.getcwd())
import pandas as pd
from datetime import datetime
import warnings
from tqdm import tqdm
import pickle
from matching.shixun.recall_comm import get_item_info_df
from deepctr.layers import custom_objects
from tensorflow.keras import backend as K
from tensorflow.keras.layers import *
from tensorflow.keras.models import *
from tensorflow.keras.callbacks import * 
from tensorflow.python.keras.models import load_model
import tensorflow as tf
from config import shixun_model_save_path
from config import logger
from config import shixun_all_user_item_feats
from config import shixun_xdeepfm_rank_dict
from config import shixun_rank_dense_fea
from config import shixun_rank_sparse_fea
from config import myshixuns_data_path
from config import shixun_features_save_path
from utils import get_user
from ranking.shixun.xdeepfm_ranker_train import get_xdeepfm_feats_columns
from ranking.shixun.user_recall_rank_features import init_rank_features
from ranking.shixun.user_recall_rank_features import build_rank_features_offline

global graph, sess

os.environ["CUDA_DEVICE_ORDER"] = "PCI_BUS_ID"
os.environ["CUDA_VISIBLE_DEVICES"] = "1"

K.set_learning_phase(True)
if tf.__version__ >= '2.0.0':
    tf.compat.v1.disable_eager_execution()
    graph = tf.compat.v1.get_default_graph()
    sess = tf.compat.v1.keras.backend.get_session()

warnings.filterwarnings('ignore')


logger.info('获取物品信息')
item_info_df = get_item_info_df()
item_info_df = item_info_df[['shixun_id', 'shixun_name']].reset_index()

logger.info('加载xDeepFM排序模型')
xdeepfm_model = load_model(shixun_model_save_path + 'xdeepfm_model.h5',custom_objects)
# xdeepfm_model.summary() 

def xdeepfm_ranker_predict(user_item_feats_df, topk=10, verbose=True):   
    """
    xDeepFM模型预测接口
    :param user_item_feats_df: 根据user_id生成的排序模型特征
    :param topk: 返回排序后的topk个物品
    """
    start_time = datetime.now()

    # 稀疏特征
    sparse_fea = shixun_rank_sparse_fea

    # 稠密连续型特征
    dense_fea = shixun_rank_dense_fea

    # 填充缺失值 
    user_item_feats_df[dense_fea] = user_item_feats_df[dense_fea].fillna(0, )

    # dense特征进行归一化
    for feat in dense_fea:
        min_max_scaler = pickle.load(open(shixun_model_save_path + 'min_max_scaler_' + feat + '.model', 'rb'))
        user_item_feats_df[feat] = min_max_scaler.transform(user_item_feats_df[[feat]])   

    # sparse特征LabelEncoder
    for feat in sparse_fea:
        label_encoder = pickle.load(open(shixun_model_save_path + feat + '_label_encoder.model', 'rb'))        
        user_item_feats_df[feat] = label_encoder.transform(user_item_feats_df[[feat]])        
        if feat == 'shixun_id':
            shixun_id_lable_encoder = label_encoder
        if feat == 'user_id':
            user_id_label_encoder = label_encoder     

    x, linear_feature_columns, dnn_feature_columns = get_xdeepfm_feats_columns(
        user_item_feats_df, dense_fea, sparse_fea)

    # 模型预测
    with graph.as_default():
        with sess.as_default():
            user_item_feats_df['pred_score'] = xdeepfm_model.predict(x, verbose=1, batch_size=256) 

    # 还原user_id和shixun_id
    user_item_feats_df['user_id'] = user_id_label_encoder.inverse_transform(user_item_feats_df[['user_id']]) 
    user_item_feats_df['shixun_id'] = shixun_id_lable_encoder.inverse_transform(user_item_feats_df[['shixun_id']])                  

    # 按预测分数降序排序
    rank_results = user_item_feats_df[['user_id', 'shixun_id', 'pred_score']]
    rank_results['user_id'] = rank_results['user_id'].astype(int)
    rank_results['shixun_id'] = rank_results['shixun_id'].astype(int)

    rank_results = rank_results.merge(item_info_df, how='left', on='shixun_id')
    rank_results = rank_results[['user_id', 'shixun_id', 'shixun_name', 'pred_score']]

    rank_results.sort_values(by=['pred_score'], ascending=False, inplace=True)
    rank_results['pred_rank'] = rank_results['pred_score'].rank(ascending=False, method='first').astype(int)
    rank_results = rank_results[:topk]

    # 计算耗时毫秒
    end_time = datetime.utcnow()
    cost_time_millisecond = round(float((end_time - start_time).microseconds / 1000.0), 3) 
    if verbose:
        logger.info(f"xDeepFM 预测耗时: {cost_time_millisecond} 毫秒")   

    return rank_results

def alluser_xdeepfm_ranker_predict():
    """
    生成所有用户召回物品离线特征排序后的字典
    """
    init_rank_features()
    all_user_item_feats_df = pd.read_csv(shixun_all_user_item_feats, sep='\t', encoding='utf-8')

    recall_rank_list_lict = {}

    # all_user_ids = all_user_item_feats_df['user_id'].unique()
    all_user_ids = get_user(myshixuns_data_path)
    for user_id in tqdm(all_user_ids):
        user_id = int(user_id)
        recall_rank_list_lict.setdefault(user_id, [])

        user_item_feats_df = build_rank_features_offline(user_id)
        if user_item_feats_df.shape[0] == 0:
            continue
        rank_results = xdeepfm_ranker_predict(user_item_feats_df, 
                                              topk=user_item_feats_df.shape[0], 
                                              verbose=False)
        
        for shixun_id, shixun_name in zip(rank_results['shixun_id'], rank_results['shixun_name']):
            recall_rank_list_lict[user_id].append((shixun_id, shixun_name))

    pickle.dump(recall_rank_list_lict, open(shixun_xdeepfm_rank_dict, 'wb'))

if __name__ == '__main__':  
    user_item_feats_df = pd.read_csv(shixun_features_save_path + 'user_item_feats_df.csv', sep='\t') 
    tmp_user_item_feats_df = user_item_feats_df.merge(item_info_df, how='left', on='shixun_id')

    logger.info('xDeepFM排序之前的数据:')    
    print(tmp_user_item_feats_df[['user_id', 'shixun_id', 'shixun_name', 'score', 'rank']][:20])

    rank_results = xdeepfm_ranker_predict(user_item_feats_df, topk=user_item_feats_df.shape[0]) 
    logger.info('xDeepFM排序之后的数据:')    
    print(rank_results[['user_id', 'shixun_id', 'shixun_name', 'pred_score', 'pred_rank']][:20])                               

    alluser_xdeepfm_ranker_predict()

    print("success!!!")