update

ad_predict_video_data_process.py

@@ -3,8 +3,11 @@ import time
 import datetime
 import pandas as pd
 from odps import ODPS
-from utils import data_check
+from utils import data_check, RedisHelper
 from threading import Timer
+from config import set_config
+config_, _ = set_config()
+redis_helper = RedisHelper()
 
 # ODPS服务配置
 odps_config = {
@@ -89,7 +92,7 @@ def video_data_process(project, table, dt, app_type):
         feature_df[column_name] = feature_df[column_name].astype(float)
     print(f"feature_df shape: {feature_df.shape}")
     print('step 3: add new video feature')
-    # 补充新用户默认数据（使用均值）
+    # 补充新视频默认数据（使用均值）
     new_video_feature = {
         'apptype': app_type,
         'videoid': '-1',
@@ -122,6 +125,29 @@ def video_data_process(project, table, dt, app_type):
     if not os.path.exists(predict_data_dir):
         os.makedirs(predict_data_dir)
     video_df.to_csv(f"{predict_data_dir}/video_feature.csv", index=False)
+    # to redis
+    xgb_config = config_.AD_MODEL_ABTEST_CONFIG['xgb']
+    for ind, row in video_df.iterrows():
+        app_type = row['apptype']
+        video_id = row['videoid']
+        value = {
+            'video_preview_count_uv_30day': row['video_preview_count_uv_30day'],
+            'video_preview_count_pv_30day': row['video_preview_count_pv_30day'],
+            'video_view_count_uv_30day': row['video_view_count_uv_30day'],
+            'video_view_count_pv_30day': row['video_view_count_pv_30day'],
+            'video_play_count_uv_30day': row['video_play_count_uv_30day'],
+            'video_play_count_pv_30day': row['video_play_count_pv_30day'],
+            'video_share_count_uv_30day': row['video_share_count_uv_30day'],
+            'video_share_count_pv_30day': row['video_share_count_pv_30day'],
+            'video_return_count_30day': row['video_return_count_30day'],
+            'video_ctr_uv_30day': row['video_ctr_uv_30day'],
+            'video_ctr_pv_30day': row['video_ctr_pv_30day'],
+            'video_share_rate_uv_30day': row['video_share_rate_uv_30day'],
+            'video_share_rate_pv_30day': row['video_share_rate_pv_30day'],
+            'video_return_rate_30day': row['video_return_rate_30day']
+        }
+        key = f"{xgb_config['predict_video_feature_key_prefix']}{app_type}:{video_id}"
+        redis_helper.set_data_to_redis(key_name=key, value=str(value), expire_time=48*3600)
 
 
 def timer_check():

config.py

@@ -1973,6 +1973,8 @@ class BaseConfig(object):
     # 广告模型abtest配置 - 模型
     AD_MODEL_ABTEST_CONFIG = {
         'xgb': {
+            # 视频特征存放 redis key 前缀，完整格式：ad:xgb:predict:video:{app_type}:{video_id}
+            'predict_video_feature_key_prefix': 'ad:xgb:predict:video:',
             # 预测结果存放 redis key 前缀，完整格式：ad:xgb:predict:{app_type}:{mid}:{video_id}
             'predict_key_prefix': 'ad:xgb:predict:',
             # 阈值计算记录存放 redis key