I'm trying to get the product_id, date and the total duplicates from a table. This table contains the id, product_id, price, wholesale_price, date_add and some more fields. Currently there are over a million of rows with products added to lookup the price history of a product. But some products have double values with the product_id and the date. For example:

So what I would like to have as output:

I have tried different things with GROUP BY, HAVING COUNT() like:

SELECT id_product, date_add, COUNT(*) as total FROM price_history
GROUP BY id_product HAVING COUNT(DISTINCT date_add) > 1;

Or:

SELECT ph.id_product, date_add, COUNT(*) AS cnt FROM price_history ph
JOIN (SELECT id_product FROM price_history HAVING COUNT(DATE_FORMAT(date_add, '%Y-%m-%d')) > 1) b
ON ph.id_product = b.id_product LIMIT 20;

CodePudding user response：

I believe that you're trying to achieve: (though I kept price in the grouping)

SELECT id_product, price, date(date_add) AS 'date', count(*) AS total 
  FROM price_history
 GROUP BY id_product, price, date(date_add) HAVING count(*) > 1;

...which gives your desired output:

id_product  price   date        total
1           6.95    2021-11-15  2
2           4.95    2021-11-16  2
1           7.95    2021-12-16  3

Example dbfiddle.