How to create a temporary table by passing in data to psycopg2? - psycopg2

Question

I have a pandas dataframe that I want to pass into a psycopg2 execute statement as a temporary table. This should be very simple:
pseudo-code...
string = """
with temporary_table (id, value) as (values %s)
select * from temporary_table
"""
cur.execute(string, df)
Where df is just a dataframe with an id and value column.
What would be the syntax to use such that I'd be able to pass this data in as a temporary table and use it in my query?

Answer 1

I would create a temporary table in Postgres database with df.to_sql or execute insert sql query with values, query it and at the end of process delete it

Answer 2

A test case that I think does what you want:
import psycopg2
from psycopg2.extras import execute_values
con = psycopg2.connect("dbname=test user=postgres host=localhost port=5432")
sql_str = """WITH temporary_table (
id,
value
) AS (
VALUES %s
)
SELECT
*
FROM
temporary_table
"""
cur = con.cursor()
execute_values(cur, sql_str, ((1, 2), (2,3)))
cur.fetchall()
[(1, 2), (2, 3)]
Using execute_values from Fast Execution Helpers.
UPDATE
Sticking to just execute:
import psycopg2
from psycopg2 import sql
con = psycopg2.connect("dbname=test user=postgres host=localhost port=5432")
cur = con.cursor()
input_data = ((1,2), (3,4), (5,6))
sql_str = sql.SQL("""WITH temporary_table (
id,
value
) AS (
VALUES {}
)
SELECT
*
FROM
temporary_table
""").format(sql.SQL(', ').join(sql.Placeholder() * len(input_data)))
cur.execute(sql_str, input_data)
cur.fetchall()
[(1, 2), (3, 4), (5, 6)]