psycopg2 returning string rather than decimal array

Question

I am querying a database with psycopg2 and python3 to get decimal arrays.

Inside the DB I am concatenating arrays with an aggregate functions like the one shown here.

From reading the psycopg2 documentation I was under the impression PostgreSQL decimal arrays should be getting converted to python decimal lists by default, but this is not happening. I am getting a single string instead.

Here is an example return value:

{"(\"{1.33,1.37,1.5,1.1,1.576...

Do I need to create a custom type-caster as outlined here on the documentation? If so, can someone point me in the right direction of how to do it because so far I am getting nowhere.

Alternative suggestions welcome. Ultimately i want these arrays in numpy objects for manipulation and plotting.

EDIT:

Ok, further digging leads me to the conclusion the issue is my aggregate function in the DB. I am using one identical to that shown in this stackoverflow question.

CREATE AGGREGATE array_accum(anyarray) (
    SFUNC = array_cat,
    STYPE = anyarray,
    INITCOND = '{}'
);

The return from this is coming as a string however.

I am calling it just like in the question linked above:

  SELECT array_accum( ARRAY[[x,x]] ) 
   FROM (SELECT subseries FROM series 
    WHERE offset BETWEEN 10 AND 20) x;

If I query for an individual subseries I am getting the correct type. After aggregation it is being converted to a string however.