Reputation: 723
I'm trying to query MySQL. Say I have 3 tables and imagine the data looked like this:
apples_sold
+--------+------------+----------+------+
| tranID | date | apple_id | sold |
+--------+------------+----------+------+
| 101 | 2012-07-01 | a01 | 2 |
| 102 | 2012-07-02 | a01 | 5 |
| 103 | 2012-07-03 | a01 | 1 |
| 104 | 2012-07-04 | a01 | 0 |
| 105 | 2012-07-05 | a01 | 2 |
+--------+------------+----------+------+
price_history
+---------+------------+----------+-------+
| priceID | date | apple_id | price |
+---------+------------+----------+-------+
| p01 | 2012-07-01 | a01 | $0.25 |
| p02 | 2012-07-03 | a01 | $0.10 | <- price change
+---------+------------+----------+-------+
apple_name
+----------+----------+
| apple_id | name |
+----------+----------+
| a01 | McIntosh |
+----------+----------+
Trying to create a query that outputs this:
apple_prices
+------------+----------+-------+
| date | name | price |
+------------+----------+-------+
| 2012-07-01 | McIntosh | $0.25 |
| 2012-07-02 | McIntosh | $0.25 |
| 2012-07-03 | McIntosh | $0.10 | <- price change
| 2012-07-04 | McIntosh | $0.10 |
| 2012-07-05 | McIntosh | $0.10 |
+------------+----------+-------+
Didn't want to store the price with the apples_sold record, normalisation and all. The problem is im unsure the best way to join the sold record with the price.
Upvotes: 0
Views: 1441
Reputation: 23135
You can use:
SELECT a.date, d.name, c.price
FROM apples_sold a
INNER JOIN
(
SELECT a.date, MAX(b.date) AS maxdate
FROM apples_sold a
INNER JOIN price_history b ON a.date >= b.date
GROUP BY a.date
) b ON a.date = b.date
INNER JOIN price_history c ON b.maxdate = c.date
INNER JOIN apple_name d ON a.apple_id = d.apple_id
Upvotes: 5