99

Lets say I have three different MySQL tables:

Table products:

id | name
 1   Product A
 2   Product B

Table partners:

id | name
 1   Partner A
 2   Partner B

Table sales:

partners_id | products_id
          1             2
          2             5
          1             5
          1             3
          1             4
          1             5
          2             2
          2             4
          2             3
          1             1

I would like to get a table with partners in the rows and products as columns. So far I was able to get an output like this:

name      | name      | COUNT( * )
Partner A   Product A          1
Partner A   Product B          1
Partner A   Product C          1
Partner A   Product D          1
Partner A   Product E          2
Partner B   Product B          1
Partner B   Product C          1
Partner B   Product D          1
Partner B   Product E          1

Using this query:

SELECT partners.name, products.name, COUNT( * ) 
FROM sales
JOIN products ON sales.products_id = products.id
JOIN partners ON sales.partners_id = partners.id
GROUP BY sales.partners_id, sales.products_id
LIMIT 0 , 30

but I would like to have instead something like:

partner_name | Product A | Product B | Product C | Product D | Product E
Partner A              1           1           1           1           2
Partner B              0           1           1           1           1

The problem is that I cannot tell how many products I will have so the column number needs to change dynamically depending on the rows in the products table.

This very good answer does not seem to work with mysql: T-SQL Pivot? Possibility of creating table columns from row values

3
  • 2
    Refer link Row to Column for multiple suggestions. Commented Feb 1, 2018 at 6:00
  • 1
    Dynamic builder Commented Jul 18, 2018 at 0:55
  • 3
    @BhavinPokiya that is an MS SQL-Server link you've provided, while this is tagged as MySQL. Commented May 12, 2020 at 17:53

1 Answer 1

127

Unfortunately MySQL does not have a PIVOT function which is basically what you are trying to do. So you will need to use an aggregate function with a CASE statement:

select pt.partner_name,
  count(case when pd.product_name = 'Product A' THEN 1 END) ProductA,
  count(case when pd.product_name = 'Product B' THEN 1 END) ProductB,
  count(case when pd.product_name = 'Product C' THEN 1 END) ProductC,
  count(case when pd.product_name = 'Product D' THEN 1 END) ProductD,
  count(case when pd.product_name = 'Product E' THEN 1 END) ProductE
from partners pt
left join sales s
  on pt.part_id = s.partner_id
left join products pd
  on s.product_id = pd.prod_id
group by pt.partner_name

See SQL Demo

Since you do not know the Products you will probably want to perform this dynamically. This can be done using prepared statements.

With dynamic pivot tables (transform rows to columns) your code would look like this:

SET @sql = NULL;
SELECT
  GROUP_CONCAT(DISTINCT
    CONCAT(
      'count(case when Product_Name = ''',
      Product_Name,
      ''' then 1 end) AS ',
      replace(Product_Name, ' ', '')
    )
  ) INTO @sql
from products;

SET @sql = CONCAT('SELECT pt.partner_name, ', @sql, ' from partners pt
left join sales s
  on pt.part_id = s.partner_id
left join products pd
  on s.product_id = pd.prod_id
group by pt.partner_name');

PREPARE stmt FROM @sql;
EXECUTE stmt;
DEALLOCATE PREPARE stmt;

See SQL Demo

It's probably worth noting that GROUP_CONCAT is by default limited to 1024 bytes. You can work around this by setting it higher for the duration of your procedure, ie. SET @@group_concat_max_len = 32000;

Sign up to request clarification or add additional context in comments.

3 Comments

if you are looking for more dynamic pivot queries please check this: boynux.com/creating-pivot-reports-in-mysql
What happens if the product name is ProductA') from partners pt; truncate partners;
@avatarofhope2 Is this a question or an implication? If this gives an injection angle what is the proper way to handle it?

Start asking to get answers

Find the answer to your question by asking.

Ask question

Explore related questions

See similar questions with these tags.