Question

Suppose I have a table with a field named ‘rating’, it may take different values, but I want to receive a count of specific values.

Example:

Create table mytable(
    rating int(1),
);

First and the obvious way I could think of was the following:

select rating,count(rating) from mytable group by rating order by rating

The problem though it is not clear how many values it would return, it may be also not easy to process them way.

What I would really like to do is to select two fields in one row showing the number of records that have some specific values.

Example…

//something like this (some pseudocode):

select count(rating=-1) as rating1, count (rating=1) as rating2 from mytable

Could you advice on some neat way I could select in the ^ above format?

Answer 1

select SUM(IF(rating=-1,1,0)) AS rating1, 
   SUM(IF(rating=1,1,0)) AS rating2 from mytable