I have a question about using subqueries in BigQuery in Datalab. This is specific to google.datalab.bigquery.Query. I can use subqueries with %bq command-line in Datalab just fine.
Assuming, I did these in Datalab by using %bq as:
%bq query --name my_table1
select col1, col2 from dataset1.table1
%bq query --subqueries my_table1
select count(col1) as some_count from my_table1 where col1 is null
Above %bq command-line codes work fine. But, I would like to use python Datalab API to do it in a more programmatic way. So in Datalab, I did:
sql_str1 = '''select col1, col2 from dataset1.table1'''
my_table1 = bq.Query(sql_str1)
sql_str2 = '''select count(col1) as some_count from my_table1 where col1 is null'''
bq.Query(sql_str2, subqueries= my_table1).execute().result()
Then, I got error messages:
TypeErrorTraceback (most recent call last)
<ipython-input-6-c625b8e326b9> in <module>()
----> 1 bq.Query(sql_command, subqueries=web_activity).execute().result()
/usr/local/lib/python2.7/dist-packages/google/datalab/bigquery/_query.pyc in __init__(self, sql, env, udfs, data_sources, subqueries)
79
80 if subqueries:
---> 81 _expand_objects(subqueries, Query, self._subqueries)
82 if udfs:
83 _expand_objects(udfs, _udf.UDF, self._udfs)
/usr/local/lib/python2.7/dist-packages/google/datalab/bigquery/_query.pyc in _expand_objects(obj_container, obj_type, target_list)
59 # and add them to the target dictionary
60 def _expand_objects(obj_container, obj_type, target_list):
---> 61 for item in obj_container:
62 # for a list of objects, we should find these objects in the given environment
63 if isinstance(obj_container, list):
TypeError: 'Query' object is not iterable
But according to document at http://googledatalab.github.io/pydatalab/google.datalab.bigquery.html, I should be able to use this:
class google.datalab.bigquery.Query(sql, env=None, udfs=None, data_sources=None, subqueries=None)
What did I do wrong? Any suggestions?