update field

src/main.py

@@ -19,15 +19,15 @@ from models.esmm.model import esmm_model_fn, model_export, model_predict
 def main():
     time_begin = time.time()
     # df = get_device_df_from_redis()
-    # df2 = get_diary_df_from_redis()
-    # print(df2.sample(1))
+    df2 = get_diary_df_from_redis()
+    print(df2.sample(1))
     # print(df.size)
     # print(df2.size)
 
     # a = device_feature_engineering(df)
     # print(a.size)
-    # b = diary_feature_engineering(df2)
-    # print(b.sample(1))
+    b = diary_feature_engineering(df2, from_redis=True)
+    print(b.sample(1))
 
     device_df, diary_df, click_df, conversion_df = read_csv_data(Path("~/data/cvr_data/"))
     print(diary_df.sample(1))

src/models/esmm/fe.py

@@ -66,7 +66,7 @@ def device_feature_engineering(df):
     return device_df[device_columns]
 
 
-def diary_feature_engineering(df):
+def diary_feature_engineering(df, from_redis=False):
     diary_df = df.copy()
     str_bool_map = {"true": True, "false": False}
 
@@ -86,9 +86,10 @@ def diary_feature_engineering(df):
     diary_df["second_positions"] = diary_df["second_positions"].apply(lambda d: d if isinstance(d, list) else [])
     diary_df["projects"] = diary_df["projects"].apply(lambda d: d if isinstance(d, list) else [])
 
-    # diary_df["is_pure_author"] = diary_df["is_pure_author"].map(str_bool_map)
-    # diary_df["is_have_pure_reply"] = diary_df["is_have_pure_reply"].map(str_bool_map)
-    # diary_df["is_have_reply"] = diary_df["is_have_reply"].map(str_bool_map)
+    if from_redis:
+        diary_df["is_pure_author"] = diary_df["is_pure_author"].map(str_bool_map)
+        diary_df["is_have_pure_reply"] = diary_df["is_have_pure_reply"].map(str_bool_map)
+        diary_df["is_have_reply"] = diary_df["is_have_reply"].map(str_bool_map)
 
     diary_df["is_pure_author"] = diary_df["is_pure_author"].astype(int)
     diary_df["is_have_pure_reply"] = diary_df["is_have_pure_reply"].astype(int)