How to I find the person who has taught the most classes - sql

Question

I want to try and find the employee who has taught the most classes as the position Teacher. So in this I want to print out Nick, as he has taught the most classes as a Teacher.
However, I am getting the error:
ERROR: column "e.name" must appear in the GROUP BY clause or be used in an aggregate function Position: 24
CREATE TABLE employees (
id integer primary key,
name text
);
CREATE TABLE positions (
id integer primary key,
name text
);
CREATE TABLE teaches (
id integer primary key,
class text,
employee integer,
position integer,
foreign key (employee) references employees(id),
foreign key (position) references positions(id)
);
INSERT INTO employees (id, name) VALUES
(1, 'Clive'), (2, 'Johnny'), (3, 'Sam'), (4, 'Nick');
INSERT INTO positions (id, name) VALUES
(1, 'Assistant'), (2, 'Teacher'), (3, 'CEO'), (4, 'Manager');
INSERT INTO teaches (id, class, employee, position) VALUES
(1, 'Dancing', 1, 1), (2, 'Gardening', 1, 2),
(3, 'Dancing', 1, 2), (4, 'Baking', 4, 2),
(5, 'Gardening', 4, 2), (6, 'Gardening', 4, 2),
(7, 'Baseball', 4, 1), (8, 'Baseball', 2, 1),
(9, 'Baseball', 4, 2);
The SQL statement I am trying to use:
SELECT count(t.class), e.name
FROM positions p
JOIN teaches t
ON p.id = t.position
JOIN employees e
ON e.id = t.employee
WHERE p.name = 'Teacher'
GROUP BY t.employee;
I've been working on this on a sql fiddle:
http://www.sqlfiddle.com/#!17/a8e19c/3

Answer 1

Your query looks pretty good. You just need to fix the GROUP BY clause, so it is consistent with the columns in the SELECT clause. Then ORDER BY and LIMIT:
SELECT count(*) cnt_classes, e.name
FROM positions p
INNER JOIN teaches t ON p.id = t.position
INNER JOIN employees e ON e.id = t.employee
WHERE p.name = 'Teacher'
GROUP BY e.id --> primary key of "employees"
ORDER BY cnt_classes DESC --> order by descending count of classes
LIMIT 1 --> keep the first row only

Answer 2

In your select you are using aggregate COUNT that counts all lines in each group (GROUP BY t.employee) but you don't aggregate e.name.
So for Nick you basically select 4 rows each for one class that have two columns - class name and teacher name. Then you ask server to count class names in Nicks group (by his employee id), that aggregates 4 rows into one with value 4 but you don't do anything about teacher name so you are left with invalid structure where you have 1 row for classes count column and 4 rows for teacher name. Same for other teachers. And that's what server is complaining about. Easiest way to fix that is to add e.name to GROUP BY, that will squeeze those 4 rows of same value into one.
To get teacher that teaches most classes you then only need to sort results by class count descending order and limit result count to 1. That will give you result row with highest class count.
Updated fiddle: http://www.sqlfiddle.com/#!17/a8e19c/7

Answer 3

You're getting the error because you need to need to have every column you're selecting (e.name in this example in the GROUP BY clause, otherwise SQL doesn't know how to group and return a count for that column. You'll also want to use TOP(1) and order by if you want to return the person with the most.
SELECT TOP(1) count(*), e.name
FROM teaches t
INNER JOIN positions p ON t.position = p.id
INNER JOIN employees e ON e.id = t.employee
WHERE p.name = 'Teacher'
GROUP BY e.name
ORDER BY count(*) DESC;