Please indicate the relationship.

And, please further explain type 1 and type 2 SCD in relation to Customer Address table

Sorry about that I've made the update.
Relationship Indication
One to many Relationships:

1. Customer to Order
   One customer can place multiple orders
2. Employee to Employee Shifts
   One employee can work multiple shifts
3. Date table to order
   Multiple orders can occur one date
   Many to Many Relationships:
4. Book Entities to Order
   Many Books can be included in multiple orders.
5. Employee to Shift
   Many employees can work multiple shifts.
   In summary, most of the relationships in the table are one to many relationships, besides
   the book entities to order and employee to shift

Excellent!

Here aggregate instead of using a true window function. it might return similar results, but not fully utilizing the windowed aggregation. Consider COUNT(...) OVER (PARTITION BY.... ORDER BY ...)

Type 1 & Type 2 in relationship to customer table

• Type 1 (Overwrite): Type 1 architecture will overwrite existing data with changes.
  When a customer moves, the address in the Customer table is simply updated. Only
  the current address is stored, and all history of past addresses is lost.
• Type 2 (History): Type 2 architecture retains changes. Each time a customer moves,
  a new row is added with the effective date of change. This way, historical addresses
  are preserved while still making it possible to identify the latest address

Thanks: SELECT customer_id, product_id, market_date, COUNT(*) OVER (PARTITION BY customer_id, product_id) AS total_customer_purchases
FROM customer_purchases;

Perfect answer!

They already multiplied by SELECT count (*) FROM customer, the CROSS JOIN is redundant.
Consider using the CROSS JOIN itself generating the Cartesian product, then sum

Thanks: SELECT customer_id, product_id, market_date, COUNT(*) OVER (PARTITION BY customer_id, product_id) AS total_customer_purchases
FROM customer_purchases;

SELECT v.vendor_name, p.product_name, sum(coalesce(v2.original_price,0) * 5) AS revenue
FROM vendor_inventory v2
JOIN vendor v ON v.vendor_id = v2.vendor_id
JOIN product p ON p.product_id = v2.product_id
CROSS JOIN customer c
GROUP BY v.vendor_name, P.product_name
ORDER BY v.vendor_name, P.product_name, revenue DESC;

Wonderful!

instead of a hardcoded timestamp, you may consider using CURRENT_TIMESTAMP

INSERT INTO product_units
VALUES (10, 'Eggs', '1 dozen', 6, 'unit', CURRENT_TIMESTAMP);

You're right!

Syntax error. Please consider
AND snapshot_timestamp = (SELECT snaphot_timestamp FROM ....)

DELETE FROM product_units
WHERE product_name = 'Eggs'
AND snapshot_timestamp = ( SELECT MIN(snapshot_timestamp)
FROM product_units
WHERE product_name ='Eggs');

Great work!