Convert sqlalchemy ORM query object to sql query for Pandas DataFrame -

this question feels fiendishly simple haven't been able find answer.

i have orm query object, say

query_obj = session.query(class1).join(class2).filter(class2.attr == 'state') 

i can read dataframe so:

testdf = pd.read_sql(query_obj.statement, query_obj.session.bind) 

but want use traditional sql query instead of orm:

with engine.connect() connection:     # execute query against database     results = connection.execute(query_obj)     # fetch results of query     fetchall = results.fetchall()     # build dataframe results     dataframe = pd.dataframe(fetchall) 

where query traditional sql string. when run error along lines of "query_obj not executable" know how convert orm query traditional query? how 1 columns in after getting dataframe?

context why i'm doing this: i've set orm layer on top of database , using query data pandas dataframe. works, it's maxing out memory. want cut in-memory overhead string folding (pass 3 outlined here: http://www.mobify.com/blog/sqlalchemy-memory-magic/). requires (and correct me if i'm wrong here) not using read_sql string , instead processing query's return raw tuples.

the long version described in detail in faq of sqlalchemy: http://sqlalchemy.readthedocs.org/en/latest/faq/sqlexpressions.html#how-do-i-render-sql-expressions-as-strings-possibly-with-bound-parameters-inlined

the short version is:

statement = query.statement print(statement.compile(engine)) 

the result of can used in read_sql.