Build multi-tenant RAG with Neon's database-per-user model — no nosy neighbors, max isolation, minimal costs

PostgreSQL SIGN() Function

Summary: in this tutorial, you will learn how to use the PostgreSQL SIGN() function to determine the sign of a number.

Introduction to the PostgreSQL SIGN() function

The SIGN() function allows you to determine the sign of a number.

Here’s the syntax of the SIGN() function:

SIGN(numeric_value)

The SIGN() function accepts a numeric value (numeric_value) and returns -1 if the value is negative, 0 if the value is zero, and 1 if the value is positive. Additionally, it returns NULL if the value is NULL.

The SIGN() function can be handy when you’re working with financial data, performing mathematical calculations, or handling data validation activities.

PostgreSQL SIGN() function examples

Let’s explore some examples of using the SIGN() function.

1) Basic SIGN() function example

The following example uses the SIGN() function to get the sign of various numbers:

SELECT
  SIGN(-10),
  SIGN(0),
  SIGN(10),
  SIGN(NULL);

Output:

sign | sign | sign | sign
------+------+------+------
   -1 |    0 |    1 | null
(1 row)

The SIGN() returns -1 for -10, 0 for 0, 1 for 10, and NULL for NULL.

2) Using the SIGN() function with table data

First, create a table called sales to store the sales amount by year:

CREATE TABLE sales (
    year INTEGER PRIMARY KEY,
    sales_amount NUMERIC NOT NULL
);

Second, insert rows into the sales table:

INSERT INTO sales (year, sales_amount)
VALUES
    (2013, 10000),
    (2014, 12000),
    (2015, 15000),
    (2016, 15000),
    (2017, 20000),
    (2018, 22000),
    (2019, 22000),
    (2020, 23000),
    (2021, 22000),
    (2022, 24000),
    (2023, 26000)
RETURNING *;

Third, compare the sales of a year with the previous year and use the SIGN() function to output the sales trend:

SELECT
    year,
    sales_amount,
    LAG(sales_amount) OVER (ORDER BY year) AS previous_year_sales,
    CASE
        WHEN LAG(sales_amount) OVER (ORDER BY year) IS NULL THEN 'N/A'
        WHEN SIGN(sales_amount - LAG(sales_amount) OVER (ORDER BY year)) = 1 THEN 'up'
        WHEN SIGN(sales_amount - LAG(sales_amount) OVER (ORDER BY year)) = -1 THEN 'down'
        ELSE 'unchanged'
    END AS sales_trend
FROM
    sales;

Output:

year | sales_amount | previous_year_sales | sales_trend
------+--------------+---------------------+-------------
 2013 |        10000 |                null | N/A
 2014 |        12000 |               10000 | up
 2015 |        15000 |               12000 | up
 2016 |        15000 |               15000 | unchanged
 2017 |        20000 |               15000 | up
 2018 |        22000 |               20000 | up
 2019 |        22000 |               22000 | unchanged
 2020 |        23000 |               22000 | up
 2021 |        22000 |               23000 | down
 2022 |        24000 |               22000 | up
 2023 |        26000 |               24000 | up
(11 rows)

How it works.

  • Use the LAG() window function to retrieve the sales_amount from the previous year utilizing the ORDER BY year clause to specify the order of rows.
  • Use the CASE expression to evaluate each row’s sales data and assign a corresponding value to the sales_trend column.

If you want to reuse the result of the LAG() function, you can use a common table expression:

WITH sales_data AS (
    SELECT
        year,
        sales_amount,
        LAG(sales_amount) OVER (ORDER BY year) AS previous_year_sales
    FROM
        sales
)
SELECT
    year,
    sales_amount,
    previous_year_sales,
    CASE
        WHEN previous_year_sales IS NULL THEN 'N/A'
        WHEN SIGN(sales_amount - previous_year_sales) = 1 THEN 'up'
        WHEN SIGN(sales_amount - previous_year_sales) = -1 THEN 'down'
        ELSE 'unchanged'
    END AS sales_trend
FROM
    sales_data;

Summary

  • Use the SIGN() function to determine the sign of a number.

Last updated on

Was this page helpful?