how can i simplify this query using partition by to find max total?

I'd like to find the max number of transactions per customer by returning the shop_id associated with that max. Below is my set of queries that gives me what i want but i am hoping to find a simpler solution.

WITH base AS (
  SELECT 
    shop_id, 
    customer_id, 
    sum(no_trans) AS total 
  FROM 
    tx.table 
  WHERE region = 'USA' 
  GROUP BY 1, 2
), 
cte2 AS (
  SELECT 
    shop_id, 
    customer_id, 
    total, 
    ROW_NUMBER() OVER (
      PARTITION BY customer_id 
      ORDER BY total DESC
    ) AS rank FROM base
), 
max_shop AS (SELECT * FROM cte2 WHERE rank = 1)

The output of cte2 looks like:

shop_id	| customer_id|total |rank
1234	| 100	     |6789	|1
345	    | 100	     |365	|2
673	    | 100	     |10	|3

and i would like only the top row returned hence rank =1 as in my max_shop cte. the above to me seems a bit long - is there a more efficient way of doing this?

Since an analytic function is evaluated later after an aggregation, you can merge base and cte2 together.

For example,

WITH tx_table AS (
  select '1234' shop_id, '100' customer_id, 6789 no_trans union all
  select '345' shop_id,  '100' customer_id, 365 no_trans union all
  select '673' shop_id,  '100' customer_id, 10 no_trans
),
cte2 AS (
  SELECT shop_id, 
         customer_id, 
         sum(no_trans) AS total,
         row_number() OVER (partition by customer_id order by sum(no_trans) desc) AS rn
    FROM tx_table
   GROUP BY shop_id, customer_id
) 
SELECT * FROM cte2 WHERE rn = 1;

Query results