SQL Spark rewrite query to not have: correlated column is not allowed in a non-equality predicate error

Question

I need almost identical query taken from here: Query for count of distinct values in a rolling date range

I wrote my own version and try to execute it in spark sql environment. However, I get the following error: Error in SQL statement: AnalysisException: Correlated column is not allowed in a non-equality predicate:

This is the query below:

SELECT
      fo.orderDate
      ,f.dist_ids
FROM  filtered_orders_date fo
LEFT JOIN LATERAL (
       SELECT
              count(DISTINCT id) as dist_ids
       FROM   filtered_orders_date
       WHERE  orderDate BETWEEN fo.orderDate - 2 AND fo.orderDate -- period of 3 days
      ) AS f ON TRUE

Is there a way to rewrite the query to get the functionality of getting distinct ids in a date range which would not give the error?