How to get the difference between the frequency of two fields? - sql

Question

I have a PostgreSQL database table that contains two columns a and b.
When I query all the entries of the table I get:
{1, 2},
{2, 3},
{2, 3}
So the value:
(1) appeared in field a 1 time and in field b 0 times
(2) appeared in field a 2 times and in field b 1 time
(3) appeared in field a 0 times and in field b 2 times
I want to get the following output:
{1, 1},
{2, 1},
{3, -2}
where the first field is the value stored in the database and the second field is difference.
How can I achieve that?
I first query the database and the result is in query_result
then I get the frequency of the first and second element:
f0 = query_result |> Enum.frequencies_by(&elem(&1, 0))
f1 = query_result |> Enum.frequencies_by(&elem(&1, 1))
(f0 |> Map.keys) ++ (f1 |> Map.keys) |> Enum.uniq |> Enum.into(%{}, fn key -> {key, (f0[key] || 0) - (f1[key] || 0)} end)
I am looking for a simpler way to do this.

Answer 1

Use a single query:
SELECT val, COALESCE(a.ct, 0) - COALESCE(b.ct, 0) AS freq_diff
FROM (
SELECT a AS val, count(*) AS ct
FROM tbl
GROUP BY 1
) a
FULL JOIN (
SELECT b AS val, count(*) AS ct
FROM tbl
GROUP BY 1
) b USING (val);
fiddle
FULL [OUTER] JOIN, because either value may be missing in the other column.
COALESCE to defend against NULL values resulting from the join.