Agenda for Class - 03/04/2014 Answer questions about HW#5 and HW#6 Review query syntax. Discuss...

Agenda for Class - 03/04/2014

Answer questions about HW#5 and HW#6

Review query syntax.

Discuss group functions and summary output with the GROUP BY statement.

Introduce sub-queries.

1

2

Structure of the SELECT statement

SELECT [all or distinct] FROM (table) WHERE (condition)GROUP BY (grouping fields)HAVING (condition)ORDER BY (sort fields)

Order of Actual Execution:1) FROM2) WHERE3) GROUP BY4) HAVING5) SELECT6) ORDER BY

Referred to as the “SELECT LIST”

When a SELECT statement is executed, the result is referred to as a “result table”. It is a memory-based table.

Table used for examples

3

Different SELECT options

4

SELECT DISTINCT deptnoFROM emp1

SELECT TOP 1 salaryFROM emp1ORDER BY salary

SELECT TOP 1 salaryFROM emp1ORDER BY salary

SELECT TOP 10 PERCENT salaryFROM emp1ORDER BY salary desc

What is a Group Function?

A way to summarize data and provide more meaningful and informative output from the database. Sometimes referred to as “summary queries” or “group functions.”

Group functions differ from single row SELECT statements:– A SELECT statement processes every row in the

underlying table. The result table (unless a WHERE clause is used) contains one row per row in the underlying table.

– A group function collects data from multiple rows and produces summarized data in the result table. There should be one row in the result table per group.

5

What do group functions produce?

If a group function is run on the whole table, without grouping, it generates a single row result table.

If a group function is run with grouping (the GROUP BY statement) then it generates one row per group in the result table.

6

Why are group functions important?

They provide information rather than data for the end-users of technology.

They help programmers understand large data sets.

7

Function Description of What is Returned

AVG Average value of a numeric column; ignores null values

COUNT Number of rows. When * is used, all rows are returned (including null values and duplicate rows)

MAX Maximum value of a column; ignores null values

MIN Minimum value of a column; ignores null values

SUM Totals the value of a numeric column; ignores null values

8

Calculating Averages

 SELECT AVG(salary)FROM emp1;

SELECT ROUND(AVG(salary),2)FROM emp1;

SELECT ROUND(AVG(salary),2)FROM emp1WHERE deptno = 10;

Use the ROUND function to perform both a

mathematical rounding operation and truncate the result to a set number of digits after the decimal

point

9

Counting Rows

SELECT COUNT(*)FROM emp1;

SELECT COUNT(*)FROM emp1WHERE deptno = 10; SELECT COUNT(*)FROM emp1WHERE salary > 2000 and deptno = 10;

SELECT COUNT(DISTINCT deptno)FROM emp1; 

10

Finding Minimum and Maximum Values

SELECT MIN(hiredate)FROM emp1;  SELECT MAX(hiredate)FROM emp1;

SELECT MIN(ename)FROM emp1;  SELECT MAX(hiredate)FROM emp1WHERE deptno = 10;

11

Groups with Calculations/Functions

SELECT MAX(salary + ISNULL(comm,0))FROM emp1;

SELECT MAX(DATEDIFF(mm, hiredate, GETDATE())FROM emp1;

12

Combining group functions

SELECT COUNT(salary), SUM(salary), MIN(salary)

FROM emp1WHERE deptno = 10 and salary < 4000;

13

Group function issue…

Combining group functions with single row values - doesn't work!!

 SELECT deptno,

COUNT(salary), SUM(salary)

FROM emp1

14

Creating summary output by grouping

SELECT deptno, SUM(salary)

FROM emp1GROUP BY deptno;

SELECT deptno, SUM(salary)

FROM emp1WHERE salary > 2000GROUP BY deptno;

Eliminates rows before the grouping

occurs

15

Summary output with conditions

SELECT deptno, SUM(salary)FROM emp1GROUP BY deptnoHAVING SUM(salary) > 6000;

 SELECT deptno, SUM(salary)FROM emp1GROUP BY deptnoHAVING AVG(salary) > 2000;  

The HAVING statement uses

group functions for the condition or

grouped attributes

16

Multi-attribute grouping

SELECT deptno, job, SUM(salary), AVG(salary)

FROM emp1GROUP BY deptno,

job;

17

A group function isn’t a condition!!

SELECT MAX(salary)FROM emp1

The statement above does identify the maximum salary from the table.

Now try to add the name of the employee to that query:

SELECT ename, MAX(salary)

FROM emp1

GROUP BY may not work exactly as you expect…

Example 1: What if you want to see the employee name of the employee who has the maximum salary?

SELECT ename,MAX(salary)

FROM emp1GROUP BY ename;

This code does not accomplish that goal!

Thoughts about fixing the problem…

SELECT ename,max(salary)

FROM emp1WHERE salary = MAX(salary)GROUP BY ename;

SELECT ename,max(salary)

FROM emp1GROUP BY enameHAVING max(salary) = salary;

They don’t work!

Could use the TOP 1 SELECT...

21

SELECT TOP 1 ename,salary,deptno

FROM emp1ORDER BY salary desc

But this works only in the SQL Server environment because TOP 1 is not

ANSI-Standard SQL

ANSI–Standard SQLNeeds a sub-query to work correctly

SELECT ename,salary,deptno

FROM emp1WHERE salary =

(SELECT MAX(salary) FROM emp1)

Example 1 answer:

What is a sub-query?

A sub-query is a query embedded inside another query.

The sub-query is executed in the normal operation of the query in which it is embedded.

The sub-query will return an “answer” result table to the query in which it is embedded.

A sub-query can be placed in the SELECT list, FROM statement, WHERE clause &/or HAVING clause.

Two types of sub-queries: non-correlated and correlated.

Types of sub-queries

Non-correlated– Inner and outer queries do not exchange data.

Correlated– Inner and outer queries do exchange data.

24

Example 2 of a non-correlated sub-query

25

Which employee has been employed with the organization for the longest time? What is the name and salary of that employee?

Or another way to look at it: What is the name and salary of the employee who was hired first in our organization

Write a query to accomplish that goal!! (Use slide #22 as your example)

Let’s expand our sample data to include a new table for department

26

27

What if we want to see all the employees who are in deptno 20? No problem...

SELECT *FROM emp1WHERE deptno = 20

Relational operators in a single table

28

What if we want to see all employees who have any department that is in the dept1

table?

SELECT *FROM emp1WHERE deptno IN

(SELECT deptno FROM dept1)

Example 3 of a non-correlated sub-query:

29

What if we want to see all employees who don’t have a department that is in the

dept1 table?

SELECT *FROM emp1WHERE deptno NOT IN

(SELECT deptno FROM dept1)

Example 4 of a non-correlated sub-query:

Example of the need for a correlated sub-query

Example 1 of a correlated sub-query: which employees have a higher salary than the average

salary for their department?SELECT deptno,AVG(salary)

FROM emp1GROUP BY deptno;

SELECT empno,ename,deptno,salary

FROM emp1ORDER BY deptno

A Correlated Sub-Query

31

SELECT empno,ename,deptno,salary

FROM emp1 outqueryWHERE salary >

(SELECT AVG(salary) FROM emp1 inquery WHERE outquery.deptno = inquery.deptno)

Requires that the tables have aliases

What if we also want to see the average salary in the SELECT list?

32

SELECT empno,ename,deptno,salary,(SELECT AVG(salary) FROM emp1 squery WHERE squery.deptno = outquery.deptno)

FROM emp1 outqueryWHERE salary >

(SELECT AVG(salary) FROM emp1 inquery WHERE outquery.deptno = inquery.deptno)