Using COALESCE to handle NULL values in PostgreSQL

Question

I have the following query

SELECT  DISTINCT 
     pt.incentive_marketing, 
     pt.incentive_channel, 
     pt.incentive_advertising 
FROM test.pricing pt 
         


        

Answer 1

You can use COALESCE in conjunction with NULLIF for a short, efficient solution:

COALESCE( NULLIF(yourField,'') , '0' )

The NULLIF function will return null if yourField is equal to the second value ('' in the example), making the COALESCE function fully working on all cases:

                 QUERY                     |                RESULT 
---------------------------------------------------------------------------------
SELECT COALESCE(NULLIF(null  ,''),'0')     |                 '0'
SELECT COALESCE(NULLIF(''    ,''),'0')     |                 '0'
SELECT COALESCE(NULLIF('foo' ,''),'0')     |                 'foo'

Answer 2

If you're using 0 and an empty string '' and null to designate undefined you've got a data problem. Just update the columns and fix your schema.

UPDATE pt.incentive_channel
SET   pt.incentive_marketing = NULL
WHERE pt.incentive_marketing = '';

UPDATE pt.incentive_channel
SET   pt.incentive_advertising = NULL
WHERE pt.incentive_marketing = '';

UPDATE pt.incentive_channel
SET   pt.incentive_channel = NULL
WHERE pt.incentive_marketing = '';

This will make joining and selecting substantially easier moving forward.