Append an element of type (composite type) on an PostgreSQL Array using Python

I have a PostgreSQL table which is defined as follow

        """ 

        CREATE TYPE similarity AS (

        user_id          integer,
        similarity       real,
        rank             integer
        )
        """,
        """
        CREATE TABLE user_similarities (

        user_id      int REFERENCES books_user_id (user_id) ON UPDATE CASCADE ON DELETE CASCADE,
        similarities similarity[]
        )
        """

to insert values I use the following function which is totally inspired from http://www.postgresqltutorial.com/postgresql-python/insert/

def insert_similarities(connection, user_id,sim_user,sim,rank):
    """ insert a similarities into user_similarities"""
    sql = """INSERT INTO user_similarities(user_id,similarities)
             VALUES(%s,%s::similarity[]);"""
    #conn = None
    #vendor_id = None
    test = [(sim_user,sim_,rank)]

    try:
        # read database configuration
        # connect to the PostgreSQL database
        conn = connection
        # create a new cursor
        cur = conn.cursor()
        # execute the INSERT statement
        cur.execute(sql, (user_id,test))
        # commit the changes to the database
        conn.commit()
        # close communication with the database
        cur.close()
    except (Exception, psycopg2.DatabaseError) as error:
        print(error)

What I want to achieve, is to append a new similarity type value to the array of similarity of a given user. In https://www.postgresql.org/docs/9.1/functions-array.html there are already existing functions that do the job but I am not using them correctly and I don't know how to do it... The function that I want to use is:

I tried a lot of sql queries, here is one an example, obviously not successful

sql = """UPDATE user_similarities SET similarities = array_append(similarities, %s::similarity[]) WHERE user_id = %s;"""

to test the output I do the following:

conn = psycopg2.connect("dbname=recommendation_t user=Etienne")
cur = conn.cursor()
cur.execute(sql,([40,0.2,2],2))

And I get:

CannotCoerce: cannot cast type integer to similarity
LINE 1: ...T similarities = array_append(similarities, ARRAY[40,0.2,2]:...
                                                           ^