Does pandas need to close connection?

Question

When using pandas "read_sql_query", do I need to close the connection? Or should I use a "with" statement? Or can I just use the following and be good?

from sqlalchemy import create_engine
import pandas as pd

sql = """
    SELECT * FROM Table_Name;
    """
engine = create_engine('blah')

df = pd.read_sql_query(sql, engine)

print df.head()

Solution 1

Looking at the source code, I can't find a con.close() method on any SQL connection object, only the cursor objects for the queries.

I'd close for safe-measure. Whether you do that using with or not is up to you.

Solution 2

For anyone who finds this question and wonders how to close the connection in this example, the following method worked for me: engine.dispose()

Comments

[juanpa.arrivillaga]

Yeah, to me is wouldn't make sense to close the connection if it is taking the connection as an argument.

[OneCricketeer]

Right, because it would need to re-open for subsequent queries.

[hoyland]

Connections should generally close when they go out of scope. That said, it's better to close connections when appropriate.

[chjortlund]

For good measure, here is the reference to official SQLAlchemy documentation with additional explanation: docs.sqlalchemy.org/en/13/core/connections.html#engine-dispo‌​sal