Lesson 4 – Using Subqueries to Solve Queries

Lesson 7 in Oracle notes

After completing this lesson, you should be able to do the following:

 Define subqueries
 Describe the type of problems that subqueries can solve
 List the types of subqueries
 Write single-row and multiple- row subqueries

This chapter covers the more advanced features of the SELECT statement.

You can write Subqueries in the where clause of another SQL statement to obtain values based on
an unknown conditional value.

This chapter covers single row subqueries and multiple row subqueries.

1
Using a Subquery to solve a problem
Problem:
Who has a salary greater than Abel’s salary?

Solution:

2 steps
 Find out how much Abel earns

 Find out who earns more than that amount

That requires two queries. We need to pass information from the first query into the second query.
Writing two separate queries does not do that.
We need a Subquery to define Abel’s salary and pass it to the main query that produces the results.

2
Subquery Syntax
A Subquery is a SELECT statement that is imbedded in a clause of another SELECT statement.

Useful when you need to select rows from a table with a condition that depend so on data from the
same table or other tables.

Where used
On the following clauses:
 WHERE clause
 HAVING clause
 FROM clause

NOTE: operator means

Single-row operator < > = etc.
Multiple-row operators IN, ANY, ALL, EXISTS

OTHER TERMS USED

Nested SELECT
Sub-SELECT
Inner SELECT

ORDER of OPERATION
The Subquery generally executes first and its output is then the fed to the main or OUTER query.

3
Using a subquery

SELECT last_name, salary

FROM employee
WHERE salary > (SELECT salary
FROM employee
WHERE last_name = 'Abel');

RESULT:
LAST_NAME SALARY
------------------------- ----------
King 24000
Kochhar 17000
De Haan 17000
Hartstein 13000
Higgins 12000

The above slide shows how we solve the problem of who earns more money than Abel.

Note that the Subquery executes first and returns the value 11,000.

The outer or main query then executes with the 11,000.

It will supply all employees with a salary greater than 11,000.

Suggestion:  … use an ORDER BY

4
Guidelines for using Subqueries:

 A Subquery must be enclosed in parenthesis.

 Place the Subquery on the right side of the comparison operator for readability
You can do it the other way
SELECT * from employees
WHERE (select salary from employee where last_name = 'Abel') < salary

 ORDER BY clause in the Subquery is only needed when performing TOP-N analysis
- Normally the order by clause is only found at the end of the SQL statement.
- TOP-N analysis refers two finding the top number of rows.
- Example top seven salaries

 2 types of Subqueries are used:

Single-row operators
Multiple-row operators

5
Types of Subqueries:

This slide shows the two types of Subqueries.

- Single-row Subqueries that return only one row from the inner SELECT statement

- Multiple-row Subqueries return more than one row from the inner SELECT statement

Special note:
There are Subqueries that return multiple columns. These are covered later.

6
 Just like other
programming
languages, the same
operators exist

Single-Row Subqueries:
For single row Subqueries that return only one row from the inner SELECT statement, single row
operators are used.

NOTE: Common Error:

You cannot use an equal to operator when you are comparing something to multiple rows.

7
PROBLEM:
Display the employees whose job ID is the same as that of employee 141

SOLUTION:
First find the job ID for employee 141

Use that job ID in the where clause to filter out the employees with the same job ID in the main
SELECT statement.

WRITE THE CODE TO DO THIS SOLUTION

Demonstrate by writing INNER query first
SELECT last_name, job_id
FROM employee
WHERE job_id = (SELECT job_id
FROM employee
WHERE employee_id = 141);

Note: I often write the inner or Subquery first to find what it returns, then I write the main query.

INNER QUERY
JOB_ID
----------
ST_CLERK

FULL RESULT
LAST_NAME JOB_ID
------------------------- ----------
Rajs ST_CLERK
Davies ST_CLERK
Matos ST_CLERK
Vargas ST_CLERK

8
 Many
subqueries
can be
used

QUERY BLOCKS
A SELECT statement is often called a query block.
 In the above example there are 3 query blocks.

The inner query block executes first bringing back the results ST_CLERK and 2600

The outer query block is then processed as if the WHERE clause was hard coded with those values
that were returned from the inner query.

NOTE:
The Subquery can get information from different tables.

9
Using Group Functions in a Subquery

PROBLEM:
Display the last name, job ID, and salary
of all employees whose salary is equal to the minimum salary of all employees.

SOLUTION:

To solve this problem

First get the minimum salary of all employees from the inner SELECT or subquery.
Secondly, use the result in the main query

 The inner query will result in a minimum salary of 2500.

 The 2500 replaces the right side of the WHERE clause

10
SELECT LAST_NAME, JOB_ID, SALARY
FROM EMPLOYEE
WHERE SALARY = (SELECT MIN (SALARY)
FROM EMPLOYEES);

Group functions in a Subquery:

This example demonstrates that you can get information from the Subquery when the Subquery has
a group function in it.

NOTE:
SELECT LAST_NAME, JOB_ID, SALARY
FROM EMPLOYEE
WHERE SALARY = MIN (SALARY);  can’t use group
function here

11
PROBLEM 1:
Display all the departments WITH minimum salary greater DEPARTMENT
50s minimum salary

Another way of saying it

Looking for all minimum salaries in each department that is greater than the minimum in department
50

12
Display all the departments WITH minimum salary greater than
DEPARTMENT 50s minimum salary

Step 1 – Find the minimum salary of department 50

- that will require a group function
SELECT min(salary) MIN(SALARY)
FROM employee -----------
WHERE department_id = 50; 2500

Step 2-Since you want to find the minimum salary in other departments you need the group function
in the main query.
SELECT department_id, min(salary)
FROM employee
GROUP BY department_id

Step 3-But you want to limit which groups are displayed. That requires a HAVING statement

Therefore the inner query is attached to the HAVING statement.

SOLUTION
SELECT department_id, min(salary)
FROM employee
GROUP BY department_id
HAVING min(salary) > ( SELECT min(salary)
FROM employee
WHERE department_id = 50);
DEPARTMENT_ID MIN(SALARY)
------------- -----------
10 4400
20 6000 Added in an ORDER BY
60 4200
80 8600
90 17000
110 8300
7000

7 rows selected

13
Using Subqueries with the HAVING clause

14
PROBLEM 2:
Find the job with the lowest average salary. Display the job ID and that
average salary.

SOLUTION: #1 Find the lowest average salary for a job ID

#2 Display that job ID and that average salary

SELECT job_id, AVG (salary)

FROM employee
GROUP BY job_id
HAVING AVG (salary) = (SELECT MIN ( AVG (salary) )
FROM employee
GROUP BY job_id );

JOB_ID AVG(SALARY)
---------- -----------
ST_CLERK 2925

What if, there are 2 job_ids with the same average salary?

15
What is wrong with This Statement?

Error:
More than one row is returned – you cannot be equal to more than one value

When you use a GROUP BY there is an implication that there will be multiple rows returned. In this
case the result of the Subquery is 7 rows returned. Each department ID in the employees table
generated a minimum salary.

The outer query cannot be equal to seven different values.

16
Solution:

SELECT department_id, employee_id, last_name, salary

Change to IN
FROM employee
WHERE salary IN (SELECT min (salary)
FROM employee
GROUP BY department_id)

DEPARTMENT_ID EMPLOYEE_ID LAST_NAME SALARY

------------- ----------- ------------------------- ----------
90 101 Kochhar 17000
90 102 De Haan 17000
60 104 Ernst 6000
60 107 Lorentz 4200
50 144 Vargas 2500
80 176 Taylor 8600
178 Grants 7000
180 de Man 7000
10 200 Whalen 4400
20 202 Fay 6000
110 206 Gietz 8300

11 rows selected

17
COMMON PROBLEM:

The above statement is correct. It didn't return any rows from the Subquery. (no Haas exists)
The query passes a null value back to the right-hand condition on the WHERE clause.
There is no job ID that is equal to NULL.
Therefore, no rows are selected

SPECIAL NOTE:
If there was a job ID with a NULL value then the left side value would be NULL, and the right-side
value would be NULL. This means that NULL would be equal to NULL and the row would be
displayed. BUT ……..

For the row to be displayed, the WHERE clause must evaluate to TRUE
Because a comparison of two NULL values results in a NULL (instead of a 1 or 0) the WHERE
condition is not true

TRY THIS ….
select last_name, department_id,salary
from employee
where department_id = ( select department_id from employee
where employee_id = 178);

Employee 178 has no department number meaning it is NULL.

Getting a little harder …

18
Multiple-Row Subqueries:

To use a Subquery that returns more than one row you need to use a Multiple-row operator

We did this before when we had a problem with the query.

We used the IN operator

SELECT department_id, employee_id, last_name, salary

FROM employee
WHERE salary IN (SELECT MIN (salary)
FROM employee
GROUP BY department_id)

DEPARTMENT_ID EMPLOYEE_ID LAST_NAME SALARY

------------- ----------- ------------------------- ----------
90 101 Kochhar 17000
90 102 De Haan 17000
60 104 Ernst 6000
60 107 Lorentz 4200
50 144 Vargas 2500
80 176 Taylor 8600
178 Grants 7000
180 de Man 7000
10 200 Whalen 4400
20 202 Fay 6000
110 206 Gietz 8300

19
Problem:
Display employees with a salary less than people with job_id of IT_PROG

Multiple-Row Subqueries:
ANY clause

Looking at the outer query, the slide displays employees who are not IT programmers
And
whose salary is less than ANY salary that is returned by the inner Subquery

The inner Subquery sends back all the salaries for JOB_ID equal to IT programmer.
The inner Subquery returns 3 salaries with values 9000, 6000 and 4200.

Since the outer query is looking for a salary less than ANY of the IT programmer salaries then it is
looking for a value that is less than 4200 and less than 6000 and less than 9000. In other words, it is
looking for a value less than the maximum value returned by the inner Subquery. The maximum
value is $9000.
This will then return IT_PROG also, unless the final line is added to the query

NOTE:
< ANY -- less than any will mean less than the maximum return
> ANY -- greater than any means more than the minimum value returned
= ANY -- equal to any is the equivalent of the IN operator

20
ALL operator

The all operator compares a value to every value

returned by a Subquery.
The example on the slide displays employees whose salary is less than the salaries of all the
employees that have a job_id of IT_PROG
AND
whose job is not the IT_PROG

Again, there are three values being returned. They are 9000, 6000 and 4200.
 To be less than ALL means you have to be less than 4200

NOTE:
> ALL -- greater than all means more than the maximum
< ALL -- less than all means less than the minimum

NOTE:

The NOT operator can be used with any of these. Caution is recommended the use of the not
operator just as it was in other programming languages.

21
PROBLEM: Display employees who do not have anyone working for them. (No subordinates)

SUBQUERY RETURNS NULL VALUES

The subquery
SELECT mgr.manager_id
FROM employee mgr
Will return 54 rows, but one of the rows is NULL

NOTE: On next page

22
NOTE: NOT IN

One of the conditions is a NULL value. The entire query returns no rows.
The problem is the NOT IN. The NOT IN is equivalent to <>ALL

All conditions that compare a NULL value returns a NULL

NOTE: IN works with NULLS

SELECT last_name
FROM employee emp
WHERE emp.employee_id IN
(SELECT mgr.manager_id
FROM employee mgr);

IN is equivalent to =ANY

NOTE:
Could have added a WHERE clause in the Subquery  WHERE manager_id is NOT NULL

SELECT last_name
FROM employee emp
WHERE emp.employee_id IN
(SELECT mgr.manager_id
FROM employee mgr);
WHERE manager_id is NOT NULL)

ASIDE:
Did we need the ALIAS table names?
No, it was done for readability

23
PRACTICE QUESTION

Prompt the user for the employee last name. The query will return last name and hire date of any
employee in the same department as the name supplied. Do not include the employee supplied.
Have use supply it

What is the INNER query?

SELECT department_id
FROM employee
WHERE last_name = ‘&Name’;

Enter ZLOTKEY and it will find nothing. Should use function UPPER on both sides

SELECT department_id
FROM employee
WHERE UPPER(last_name) = UPPER('&Name');

Now do the outer query

SELECT last_name, department_id

FROM employee
WHERE department_id = ( SELECT department_id
FROM employee
WHERE UPPER(last_name) = UPPER('&Name') )

Now eliminate the name entered

SELECT last_name, department_id
FROM employee
WHERE department_id = ( SELECT department_id
FROM employee
WHERE UPPER(last_name) = UPPER('&&Name') )
AND UPPER (last_name) < > UPPER ('&Name');

UNDEFINE NAME;

24
Multiple Column Sub Query

A multiple-column subquery returns more than one column to the outer query and can be listed in the
outer query's FROM, WHERE, or HAVING clause. For example, the below query shows the
employee or employees in each department whose current salary is the lowest (or minimum) salary in
the department.

SELECT last_name, department_id, salary

FROM employee
WHERE (department_id, salary) IN (SELECT department_id, min(salary)
FROM employee
GROUP BY department_id)
ORDER BY department_id

The sub query returns the following:

DEPARTMENT_ID MIN(SALARY)
------------- -----------
10 4400
20 6000
50 2500
60 4200
80 8600
90 17000
110 8300
7000
8 rows selected

The full query returns

LAST_NAME DEPARTMENT_ID SALARY
------------------------- ------------- ----------
Whalen 10 4400
Fay 20 6000
Vargas 50 2500
Lorentz 60 4200
Taylor 80 8600
Kochhar 90 17000
De Haan 90 17000
Gietz 110 8300

8 rows selected

NOTE: In department 90 is 2 people with the same minimum. Since both the sub and the full query
returned 8 rows, then there must be a row missing in the full query.

 The NULL department did not show.

How would you fix this? Assuming the user wants to show the results where there is no department

25
26