add test file

tensnsorflow/multi.py

@@ -297,7 +297,7 @@ def get_predict(date,value_map,app_list_map,leve2_map,leve3_map):
           "left join jerry_test.cart_tag cart on e.device_id = cart.device_id " \
           "left join jerry_test.knowledge k on feat.level2 = k.level2_id " \
           "left join jerry_test.search_doris doris on e.device_id = doris.device_id and e.stat_date = doris.get_date " \
-          "limit 100000"
+          "limit 60000"
 
     features = ["ucity_id", "ccity_name", "device_type", "manufacturer",
                 "channel", "top", "time", "hospital_id",
@@ -342,9 +342,9 @@ def get_predict(date,value_map,app_list_map,leve2_map,leve3_map):
     # native_pre.toPandas().to_csv(local_path+"native.csv", header=True)
     spark.createDataFrame(rdd.filter(lambda x: x[0] == 0)
                           .map(lambda x: (x[1], x[2], x[6], x[7], x[8], x[9], x[10], x[11],
-                                          x[12], x[13], x[14], x[15], x[16], x[17], x[18]))) \
+                                          x[12], x[13], x[14], x[15], x[16], x[17], x[18],x[3],x[4],x[5]))) \
         .toDF("y", "z", "app_list", "level2_list", "level3_list", "tag1_list", "tag2_list", "tag3_list", "tag4_list",
-              "tag5_list", "tag6_list", "tag7_list", "ids", "search_tag2_list","search_tag3_list") \
+              "tag5_list", "tag6_list", "tag7_list", "ids", "search_tag2_list","search_tag3_list","city","uid","cid_id") \
         .repartition(1).write.format("tfrecords").save(path=path+"test_native/", mode="overwrite")
     print("native tfrecord done")
     h = time.time()
@@ -357,9 +357,9 @@ def get_predict(date,value_map,app_list_map,leve2_map,leve3_map):
 
     spark.createDataFrame(rdd.filter(lambda x: x[0] == 1)
                           .map(lambda x: (x[1], x[2], x[6], x[7], x[8], x[9], x[10], x[11],
-                                          x[12], x[13], x[14], x[15], x[16], x[17], x[18]))) \
+                                          x[12], x[13], x[14], x[15], x[16], x[17], x[18],x[3],x[4],x[5]))) \
         .toDF("y", "z", "app_list", "level2_list", "level3_list", "tag1_list", "tag2_list", "tag3_list", "tag4_list",
-              "tag5_list", "tag6_list", "tag7_list", "ids", "search_tag2_list","search_tag3_list") \
+              "tag5_list", "tag6_list", "tag7_list", "ids", "search_tag2_list","search_tag3_list","city","uid","cid_id") \
         .repartition(1).write.format("tfrecords").save(path=path + "test_nearby/", mode="overwrite")
     print("nearby tfrecord done")