Oracle FIRST_VALUE function

Oracle/PLSQL FIRST_VALUE function returns the first value in an ordered set of values from the analytical window. It is similar to FIRST_VALUE and NTH_VALUE.

Oracle/PLSQL syntax of FIRST_VALUE function

FIRST_VALUE (_id)
[RESPECT NULLS | IGNORE NULLS]
OVER ([query_partition_clause_id] [order_by_clause_id])

The following syntax is also the accepted format:

FIRST_VALUE (_id [RESPECT NULLS | IGNORE NULLS])
OVER ([query_partition_clause_id] [order_by_clause_id])

where:

• expression_id – The column or expression for which you want to return the first value.
• RESPECT NULLS | IGNORE NULLS – Optional. It determines whether NULL values are enabled or ignored in the analysis window. If this parameter is omitted, the default value is RESPECT NULLS, which includes NULL values.
• query_partition_clause_id – Optional. It is used to divide results into groups based on one or more expressions.
• order_by_clause_id – Optional. It is used to organize the data in each section.
• windowing_clause_id – Optional. It defines the rows in the analysis window for evaluation, and it is important that you use the correct windowing_clause, or you may get unexpected results. This can be a value such as:

• FIRST_VALUE returns the first value represented by its data type.

You can use FIRST_VALUE in the following versions of Oracle/PLSQL

Oracle 12c, Oracle 11g, Oracle 10g, Oracle 9i

DDL / DML for examples

If you want to follow this guide, use the DDL to create the table and DML to fill in the data. Then try the examples in your own database!

DDL / DML for FIRST_VALUE

If you have an Oracle database and want to explore the capabilities of the FIRST_VALUE analytics function yourself, then below we provide the DDL and DML that you will need.

Just follow the instructions to complete your database. Then follow the link to return to the FIRST_VALUE training material to try out examples for yourself.

DDL Example

DDL means the data definition language and is the instruction required to create the tables used in the FIRST_VALUE example.

Follow the following DDL instructions in your Oracle database:

CREATE TABLE empls
( empl_id number(6) NOT NULL,
f_name varchar2(25) NOT NULL,
l_name varchar2(30) NOT NULL,
salary_id number(8,2),
depart_id number(4),
CONSTRAINT emp_id_pk1 PRIMARY KEY (empl_id);

DML example

DML means a data manipulation language. These are INSERT commands that you will need to run in your Oracle database to populate the data:

Run the following DML operators in your Oracle database:

insert into empls (empl_id, f_name, l_name, salary_id, depart_id)
(100, 'Anita', 'Borg', 2500.10);

insert into empls (empl_id, f_name, l_name, salary_id, depart_id)
(200, "Alfred", "Aho", 3200.10);

insert into empls (empl_id, f_name, l_name, salary_id, depart_id)
(300, "Bill", "Gates", 2100.10);

insert into empls (empl_id, f_name, l_name, salary_id, depart_id)
(400, 'Linus', 'Torvalds', 3700.20);

insert into empls (empl_id, f_name, l_name, salary_id, depart_id)
(500, "Michael", "Dell", 3100.20);

insert into empls (empl_id, f_name, l_name, salary_id, depart_id)
(600, "Nello", "Cristianini", 2950, 20);

insert into empls (empl_id, f_name, l_name, salary_id, depart_id)
(700, "Rasmus", "Lerdorf", 4900.20);

insert into empls (empl_id, f_name, l_name, salary_id, depart_id)
(800, "Steve", "Jobs", 2600.30);

insert into empls (empl_id, f_name, l_name, salary_id, depart_id)
900, 'Thomas', 'Kyte', 5000.30;

Example

Let’s look at some examples of Oracle FIRST_VALUE functions and see how to use FIRST_VALUE in Oracle / PLSQL.

The highest salary for all employees
Start with a simple example and use the FIRST_VALUE function to return the highest salary in the employee table. In this example we do not need to query_partition_clause because we are evaluating the entire employees table.

In this example we have a table of employees with the following data:

To find the highest salary, enter the following SELECT:

SELECT DISTINCT FIRST_VALUE(salary_id)
OVER (ORDER BY salary_id DESC)
RANGE BETWEEN UNBOUNDED PRECEDING AND UNBOUNDED FOLLOWING)
AS "HIGH"
FROM empls;

The result you should get:

In this example, FIRST_VALUE returns the highest salary value specified by FIRST_VALUE (salary_id). The analysis window sorts the salary data in descending order, as specified by ORDER BY salary DESC. The parameter windowing_clause = RANGE BETWEEN UNBOUNDED PRECEDING AND UNBOUNDED FOLLOWING is used to ensure that all rows are included regardless of the current row.

And since we want to get the highest salary in the table, we did not need to enable query_partition_clause_id to separate the data.

The highest salary for depart_id
Now let’s show you how to use FIRST_VALUE with query_partition_clause_id. In the following example, let’s return the highest salary for depart_id 10 and 20.

Based on the same table, enter the following SQL statement:

SELECT DISTINCT depart_id, FIRST_VALUE(salary_id)
OVER (PARTITION BY depart_id ORDER BY salary_id DESC)
RANGE BETWEEN UNBOUNDED PRECEDING AND UNBOUNDED FOLLOWING)
AS "HIGH"
FROM empls
WHERE depart_id in (10,20)
ORDER BY depart_id;

Here are the results you should get:

In this example, FIRST_VALUE returns the highest salary value specified by FIRST_VALUE (salary_id). The analysis window will divide the results by DEPARTMENT_ID and organize the salary data in descending order, as specified by PARTITION BY DEPARTMENT_ID ORDER BY DESC salary.

Lowest salary by department_id
Now let’s show you how to use the FIRST_VALUE function to return the lowest salaries for department_id 10 and 20.

Again, based on the data in the table, enter the following SQL statement:

SELECT DISTINCT depart_id, FIRST_VALUE(salary_id)
OVER (PARTITION BY depart_id ORDER BY salary_id ASC)
RANGE BETWEEN UNBOUNDED PRECEDING AND UNBOUNDED FOLLOWING)
AS "LOW"
FROM empls
WHERE depart_id in (10,20)
ORDER BY depart_id;

Here are the results you should get:

In this example, we have changed the sorting order by section as PARTITION BY DEPART_ID ORDER BY salary ASC, and now we get the lowest salary based on DEPART_ID.

FIRST_VALUE and LAST_VALUE : Problem Solving using Analytic Functions