CODEWITHSUNDEEP
sqlpostgresqlstatistics
588chm
588chm

Reputation: 227

How to create "statistics" for every table of a schema? (where statistics means 1 returned record per each table)

I have a schema myschema in which I have dozens of tables, each containing a specific column maincolumn. I would like to create a statistics of the whole schema, where I can see how many distinct values are in the maincolumn per each table.

I know I can list the tables in a schema as a result of this query:

SELECT tablename 
FROM pg_tables 
WHERE schemaname = 'myschema'

Results:

tablename
---------
table1
table2
table3
...

Also I can run the query for each table:

SELECT COUNT(DISTINCT(maincolumn))  
FROM myschema.table1

Results:

count
-----
2972

But I wonder if there's a way to join these two tasks into one query, which would result this:

tablename | count
----------+------
table1    | 2972 
table2    |  542 
table3    | 1523 
....

A not so elegant solution would be to list all my tables, put the result in an Excel table and use CONCATENATE function to create a SELECT query to every table, and then I could run it line by line, but I would prefer an easier way.

Upvotes: 1

Views: 286

Answers (1)

Laurenz Albe
Laurenz Albe

Reputation: 246393

You would need dynamic SQL for that, probably in a PL/pgSQL function like this:

CREATE OR REPLACE FUNCTION table_counts() RETURNS TABLE (table_name text, count bigint)
   LANGUAGE plpgsql AS
$$BEGIN
   FOR table_counts.table_name IN
      SELECT t.table_name
      FROM information_schema.tables AS t
      WHERE t.table_schema = 'myschema'
        AND t.table_type = 'BASE TABLE'
   LOOP
      EXECUTE format(
                 'SELECT count(*) FROM myschema.%I',
                 table_counts.table_name
              ) INTO table_counts.count;

      RETURN NEXT;
   END LOOP;
END;$$;

Also, this is quite expensive.

If approximate counts are good enough for you, try

SELECT c.relname, c.reltuples
FROM pg_catalog.pg_class AS c
   JOIN pg_catalog.pg_namespace AS n
      ON c.relnamespace = n.oid
WHERE n.nspname = 'myschema'
  AND c.relkind = 'r';

The result becomes more accurate if you run ANALYZE first.

Upvotes: 2

Related Questions