## Having
`WHERE` cannot work with aggregates, `HAVING` can.
```sql
SELECT column_name(s)
FROM table_name
WHERE condition
GROUP BY column_name(s)
HAVING condition
ORDER BY column_name(s);
```
(see: [W3Schools](https://www.w3schools.com/sql/sql_having.asp) and [this stackoverflow](https://stackoverflow.com/questions/301793/mysql-using-count-in-the-where-clause)).

So e.g. if we wanted, naively, to do
```sql
SELECT DISTINCT gid
FROM `gd`
WHERE COUNT(*) > 10            --- error
ORDER BY lastupdated DESC
```
this wouldn't work. This is where we need `HAVING`:
```sql
SELECT gid
FROM `gd`
GROUP BY gid 
HAVING COUNT(*) > 10
ORDER BY lastupdated DESC
```
Essentially, `WHERE` happens before aggregation, whereas `HAVING` happens after aggregation.
See *Order of Execution* below.

# Order of Execution
(See [builtin.com](https://builtin.com/data-science/sql-order-of-execution)).

These clauses are *not* evaluated in the order written, but rather in the order below.
Importantly, column aliases created in the `SELECT` cannot be referred to by the `WHERE`
and `HAVING` clauses.

1. `FROM/JOIN`
1. `WHERE`
1. `GROUP BY`
1. `HAVING`
1. `SELECT`
1. `ORDER BY`
1. `LIMIT/OFFSET`