按照用户去重应用列表样本

tensnsorflow/applist.py

@@ -29,10 +29,13 @@ def json_format(x):
 
 def sort_app():
     db = pymysql.connect(host='10.66.157.22', port=4000, user='root', passwd='3SYz54LS9#^9sBvC', db='jerry_prod')
-    sql = "select device_id,app_list from device_id_applist"
+    sql = "select device_id,app_list,stat_date from device_id_applist"
     df = con_sql(db, sql).dropna()
-    df = df.rename(columns={0: "device_id", 1: "app_list"})
+    df = df.rename(columns={0: "device_id", 1: "app_list",2:"stat_date"})
     print(df.shape)
+    df = df.sort_values(by="stat_date",ascending=False)
+    print(df.head())
+    df = df.drop("stat_date",axis=1)
     df = df.drop_duplicates("device_id")
     print(df.shape)
     df = df.loc[df["app_list"].apply(is_json)]

tensnsorflow/ffm.py

@@ -177,6 +177,7 @@ def get_data():
     for i in ["ucity_id","clevel1_id","ccity_name","device_type","manufacturer","channel","top"]:
         features = features + len(df[i].unique())
         df[i] = df[i].fillna(0)
+    df["time"] = df["time"].fillna(df["time"].mode()[0])
     print(df.count())
     print(df.head(2))
     print("fields:{}".format(df.shape[1]-1))