Finding aggregate data size in columns

Hi,

We have a business management product that we sell to small businesses.
Over the last few months we noticed that some of the backups are growing in
size much faster than they should. Obviously I am persisting something that
I should not.

After consulting Dr Goggle I have a simple query to determine candidate
culprit tables that are growing too fast. (see below). It includes TOAST
data and it has identified a few candidate tables that I would never have
considered investigating.

Now I have some candidate tables - but I'm still struggling to identify
which are the culprit columns within those tables that are taking up the
space.

Can anyone advise how to get the aggregate data size in columns (including
toasted data)?

So I am looking for column data size (inc toast) PER TABLE (not per row).

In the scheme of things these databases are small (generally << 100MB on
disk), so elegance, efficiency & performance are not significant issues.

Any suggestions, guidance, tips, tricks etc would be much appreciated.

Many thanks for your time,
-Damian

==== Query to check data usage of tables ============

SELECT relname AS "relation_my_table_name",

pg_size_pretty(pg_total_relation_size(C.oid)) AS "total_size_inc_toast",

pg_size_pretty(relpages::bigint*8*1024) AS "size_sans_toast",

pg_size_pretty((pg_total_relation_size(C.oid)) -
(relpages::bigint*8*1024)) AS "toasted_size",

reltuples AS "#_of_entries",

to_char((pg_total_relation_size(C.oid)/reltuples),'999999D9') AS
"size_per_entry"

FROM pg_class C

LEFT JOIN pg_namespace N ON (N.oid = C.relnamespace)

WHERE nspname NOT IN ('pg_catalog', 'information_schema')

AND C.relkind <> 'i'

AND nspname !~ '^pg_toast'

AND reltuples > 0 -- avoid DivByZero in size_per_entry

ORDER BY pg_total_relation_size(C.oid) DESC

LIMIT 20;