Best way to query a table with a predefined macro in DuckDB

[yannsartori]

Hello all,

I've decided that Ibis would be a great tool for me to work with DuckDB files. Currently, I produce a .ddb file and hand it off to another person to query.

An example dataset I have in this .ddb file I produce is a bunch of temperatures e.g.

import ibis
import pandas as pd
data = pd.DataFrame({"temperature": [0,0,1,5,3], "count": [0,1,2,3,4]})
conn = ibis.duckdb.connect()
temperatures = conn.create_table("temperatures", data)

Since temperature values can exist in a bunch of tables, I want to use a macro to make it easier to query. It makes sense to me, given the nature of ibis, that there is no nice API to create it, so I resort to raw_sql:

conn.raw_sql("CREATE OR REPLACE MACRO celsius_to_fahrenheit(temp) AS 32 + (temp * (9/5))")

However, I couldn't find a nice way for the end-user to query temperatures with my macro. The best I have is:

conn.sql("SELECT celsius_to_fahrenheit(temperature) FROM temperatures").execute()

which to me, defeats the purpose of using Ibis. I would like for something slightly more developer-friendly, something akin to maybe:

temperatures.select(celsius=temperatures.temperature, fahrenheit=temperatures.sql("celsius_to_farenheit(temperature)"))

but that gives me an exception. Is there any better way of utilizing it than using connection.sql?