修改测试文件

tensnsorflow/multi.py

@@ -34,7 +34,7 @@ def feature_engineer():
     validate_date = con_sql(db, sql)[0].values.tolist()[0]
     print("validate_date:" + validate_date)
     temp = datetime.datetime.strptime(validate_date, "%Y-%m-%d")
-    start = (temp - datetime.timedelta(days=300)).strftime("%Y-%m-%d")
+    start = (temp - datetime.timedelta(days=6)).strftime("%Y-%m-%d")
     print(start)
 
     sparkConf = SparkConf().set("spark.hive.mapred.supports.subdirectories", "true") \
@@ -80,9 +80,23 @@ def feature_engineer():
                               "channel", "top", "time", "stat_date", "app_list", "hospital_id", "level3_ids"])
 
     df = df.fillna("na")
-    df = df.toPandas()
-    print(df.shape)
-    print(df.head(6))
+
+    app_list_value = [i.split(",") for i in df.select("app_list").collect().unique()]
+    app_list_unique = []
+    for i in app_list_value:
+        app_list_unique.extend(i)
+    app_list_unique = list(set(app_list_unique))
+    number = len(app_list_unique)
+    app_list_map = dict(zip(app_list_unique, list(range(1, number + 1))))
+
+    df = df.select("app_list","ucity_id", "level2_ids", "ccity_name", "device_type", "manufacturer","channel",
+                   "top", "time", "stat_date", "hospital_id", "level3_ids","y","z",
+                   "treatment_method","price_min","price_max","treatment_time","maintain_time","recover_time")\
+        .map(lambda x :app_list_func(x[0],app_list_map))
+
+    df.show(6)
+
+