Great work! We can also create another version of this report by changing revenue values into revenue deltas. The end result should look like this:

We can use the following query:

SELECT 
  EXTRACT(year FROM order_date) AS revenue_year,
  SUM(oi.amount) - LAG(SUM(oi.amount)) OVER (ORDER BY EXTRACT(year FROM order_date)) AS total_revenue_delta, 
  SUM(CASE WHEN c.category_name = 'Beverages' THEN oi.amount ELSE 0.0 END)
  - LAG(SUM(CASE WHEN c.category_name = 'Beverages' THEN oi.amount ELSE 0.0 END)) OVER (ORDER BY EXTRACT(year FROM order_date)) AS beverages_revenue_delta,
  SUM(CASE WHEN c.category_name = 'Seafood' THEN oi.amount ELSE 0.0 END)
  - LAG(SUM(CASE WHEN c.category_name = 'Seafood' THEN oi.amount ELSE 0.0 END)) OVER (ORDER BY EXTRACT(year FROM order_date)) AS seafood_revenue_delta 
FROM order_items oi 
JOIN orders o
  ON oi.order_id = o.order_id
JOIN products p
  ON p.product_id = oi.product_id
JOIN categories c
  ON c.category_id = p.category_id
GROUP BY EXTRACT(year FROM order_date)
ORDER BY EXTRACT(year FROM order_date);

We now calculate each column as the sum from the current year (SUM(oi.amount)) minus the sum from the previous year (LAG(SUM(oi.amount)) OVER (ORDER BY EXTRACT(year FROM order_date))). To that end, we used the LAG() OVER() pattern we learned in the previous part.