解决过滤前样本是0的bug

diaryCandidateSet.py

@@ -9,9 +9,10 @@ def filter_cid(df):
     data_set_cid = pd.read_csv(DIRECTORY_PATH + "data_set_cid.csv")["cid"].values.tolist()
     print("过滤前样本大小:")
     print(df.shape)
-    df = df.loc[df["cid"].isin(data_set_cid)]
-    print("过滤后样本大小:")
-    print(df.shape)
+    if not df.empty:
+        df = df.loc[df["cid"].isin(data_set_cid)]
+        print("过滤后样本大小:")
+        print(df.shape)
     return df
 
 

processData.py

@@ -101,6 +101,6 @@ def ffm_transform(data, test_number, validation_number):
 
 if __name__ == "__main__":
     data_fe = feature_en()
-    # ffm_transform(data_fe)
+    ffm_transform(data_fe)
 
 

userDiaryPredict.py

@@ -10,18 +10,21 @@ from userProfile import fetch_user_profile
 
 
 # 将device_id、city_id拼接到对应的城市热门日记表。注意：下面预测集特征顺序要与训练集保持一致
-def device_id_merge(user_profile):
+def feature_en(user_profile):
     file_name = DIRECTORY_PATH + "diaryTestSet/{0}DiaryTop2000.csv".format(user_profile['city_id'])
     data = pd.read_csv(file_name)
 
     data["device_id"] = user_profile['device_id']
+
     now = datetime.datetime.now()
     data["hour"] = now.hour
     data["minute"] = now.minute
+
     data.loc[data["hour"] == 0, ["hour"]] = 24
     data.loc[data["minute"] == 0, ["minute"]] = 60
     data["hour"] = data["hour"].astype("category")
     data["minute"] = data["minute"].astype("category")
+
     data["y"] = 0
     data = data.drop("city_id", axis=1)
     print(data.head(2))
@@ -37,7 +40,7 @@ def transform_ffm_format(df, device_id):
         data = ffm_format_pandas.transform(df)
         now = datetime.datetime.now().strftime("%Y-%m-%d-%H-%M")
 
-        predict_file_name = DIRECTORY_PATH + "diaryPredictSet/{0}_{1}DiaryTop2000.csv".format(device_id, now)
+        predict_file_name = DIRECTORY_PATH + "result/{0}_{1}DiaryTop2000.csv".format(device_id, now)
         data.to_csv(predict_file_name)
         user_instance_file_path = ''
         return user_instance_file_path
@@ -45,15 +48,33 @@ def transform_ffm_format(df, device_id):
 
 # 将模型加载，预测，把预测日记的概率值按照降序排序，存到一个表里
 def predict(user_profile):
-    ffm_model = xl.create_ffm()
-
-    user_instance = device_id_merge(user_profile)
+    user_instance = feature_en(user_profile)
     user_instance_file_path = transform_ffm_format(user_instance)
 
+    ffm_model = xl.create_ffm()
     ffm_model.setTest(user_instance_file_path)
     ffm_model.predict(DIRECTORY_PATH + "model_{0}-{1}_lr{2}_lambda{3}.out".format(DATA_START_DATE,
                                                                                   DATA_END_DATE, lr, l2_lambda),
-                      DIRECTORY_PATH + "/{0}_output.txt".format(user_profile['device_id']))
+                      DIRECTORY_PATH + "result/{0}_output.txt".format(user_profile['device_id']))
+
+    upload_predict(user_profile, user_instance)
+
+
+def upload_predict(user_profile, instance):
+    probabilities = pd.read_csv(DIRECTORY_PATH +
+                                "result/{0}_output.txt".format(user_profile['device_id']), header=None)
+    probabilities = probabilities.rename(columns={0: "prob"})
+    probabilities["cid"] = instance['cid']
+    probabilities = probabilities.sort_values(by="0",ascending=False)
+
+    wrapper_result(probabilities,user_profile['device_id'])
+
+
+def wrapper_result(prob,device_id):
+    prob = prob.head(500)
+    prob["url"] = prob["cid"].apply(lambda x:"http://m.igengmei.com/diary_book/" + prob[x.index('|') + 1:] + '/')
+    prob.to_csv(DIRECTORY_PATH+"result/{}_feed".format(device_id))
+
 
 
 def router(device_id):
@@ -61,8 +82,9 @@ def router(device_id):
 
     if is_exist:
         predict(user_profile)
+
     else:
-        print('Sorry, we don\'t have you')
+        print('Sorry, we don\'t have you.')
 
 
 if __name__ == "__main__":

userProfile.py

@@ -5,5 +5,4 @@ def fetch_user_profile(device_id):
     # TODO sql语句中的device_id可能对应多个city_id
     sql = "select device_id,city_id from data_feed_click limit 1"
     user_profile = con_sql(sql)
-    is_exist = user_profile.empty
-    return user_profile, is_exist
+    return user_profile, user_profile.empty