fix problems with automl docs (#2501)

Today when we try to use the function `batch_predict` follow the docs we receive and error saying: `the paramaters should be a pandas.Dataframe` it’s happens because the first parameter of the function `batch_predict` is a pandas.Dataframe. To solve this problem we need to use de named parameters of python.

Author: alefhsousa

tables/automl/automl_tables_predict.py

@@ -72,9 +72,9 @@ def batch_predict(project_id,
     client = automl.TablesClient(project=project_id, region=compute_region)
 
     # Query model
-    response = client.batch_predict(
-        gcs_input_uris, gcs_output_uri, model_display_name=model_display_name
-    )
+    response = client.batch_predict(gcs_input_uris=gcs_input_uris,
+                                    gcs_output_uri_prefix=gcs_output_uri,
+                                    model_display_name=model_display_name)
     print("Making batch prediction... ")
     response.result()
     print("Batch prediction complete.\n{}".format(response.metadata))