修改测试文件

tensnsorflow/multi.py

@@ -80,22 +80,25 @@ def feature_engineer():
                               "channel", "top", "time", "stat_date", "app_list", "hospital_id", "level3_ids"])
 
     df = df.fillna("na")
+    v = df.select("app_list").collect()
+    print(type(v))
+    print(v[:2])
 
-    app_list_value = [i.split(",") for i in set(df.select("app_list").collect())]
-
-    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)
+    # app_list_value = [i.split(",") for i in v]
+    #
+    # 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)