Alias not working with multiple tables

[quasiben]

Alias not working with multiple tables

[alexander-beedie]

Aliasing works fine, but we don't (currently) support the old-style "implicit join" syntax; using the more modern/standard JOIN … ON and/or JOIN … USING syntax will execute without any trouble:

ldf = pl.DataFrame({
    "a": [1, 2, 3, 4, 5, 6, 7, 1, 1],
    "b": [1, 1, 1, 1, 1, 1, 1, 10, 20],
})
rdf = pl.DataFrame({
    "a": [1, 2, 3, 4, 5, 6, 7],
    "b": [1, 1, 1, 1, 1, 1, 1],
})

pl.sql(
    """
    SELECT
        SUM(l.a) AS la_total,
        SUM(l.b) AS lb_total,
        SUM(r.b) as rb_total
    FROM
        ldf l JOIN rdf r USING (a, b)
    """
).collect()

# shape: (1, 3)
# ┌──────────┬──────────┬──────────┐
# │ la_total ┆ lb_total ┆ rb_total │
# │ ---      ┆ ---      ┆ ---      │
# │ i64      ┆ i64      ┆ i64      │
# ╞══════════╪══════════╪══════════╡
# │ 28       ┆ 7        ┆ 7        │
# └──────────┴──────────┴──────────┘

Now, I am looking at adding support for implicit joins, but I'd recommend use of JOIN … ON and/or JOIN … USING for SQL queries generally (and for the time-being Polars requires it) 👍

We have an open issue for implicit join support here: #16662.

[quasiben]

Thanks for the quick reponse @alexander-beedie . In the example above there is no join just a select

[alexander-beedie]

Thanks for the quick reponse @alexander-beedie . In the example above there is no join just a select

There is - what you have written is the old-school syntax for an implicit Cartesian join of the two tables (which is possibly not what you intended?) ;)

Do you actually want to union the two tables? (You have to relate them somehow)

[quasiben]

I don't actually want the union. I was mostly playing around with sql and was trying to naively read/select. Thank you