Find the count of EMPTY or NULL columns in a MySQL table

Question

I have around 30 columns in a MySQL table. I want to calculate how many column fields for a particular row are empty. This table is for storing user information. I want to fin

Answer 1

If the idea is to get a result something like this:

col       emp    
------    ------    
FName        15  
LName         2  
Age          22 

..use:

SELECT 'FName' AS col, SUM(CASE FName IS NULL || FName='' THEN 1 ELSE 0 END) as emp FROM MyTable
UNION
SELECT 'LName' AS col, SUM(CASE LName IS NULL || LName='' THEN 1 ELSE 0 END) as emp FROM MyTable
UNION
SELECT 'Age' AS col, SUM(CASE Age IS NULL || Age='' THEN 1 ELSE 0 END) as emp FROM MyTable

...or:

SELECT SUM(CASE t.fname IS NULL OR t.fname = '' THEN 1 ELSE 0 END) AS fname_count,
       SUM(CASE t.lname IS NULL OR t.lname = '' THEN 1 ELSE 0 END) AS lname_count,
       SUM(CASE t.age IS NULL OR t.age = '' THEN 1 ELSE 0 END) AS age_count
  FROM MYTABLE t

Answer 2

You can compute it like this:

SELECT SUM((`Name` = '') + (`Age` = 0) + (`Location` = '' OR `Location` IS NULL) + ...)

You didn't specify what types of columns you have so I used different comparing methods to illustrate the point. As a general idea you should compare to see if a specific column value is equal to the default value for that field (i.e. field not specified by the user). Use = '' for strings, = 0 for numbers, IS NULL for columns that have default NULL etc. You could also combine checks for NULL and empty value if you want. It all depends on what you want to find out.

Of course, as dnagirl pointed out in her answer, you should test for both empty and NULL values, but that really depends on how your columns are defined and also if you consider empty values to be filled or not. The point I was trying to make is that you can use SUM to add up boolean expression results.

EDIT: Didn't notice that you wanted for each row. Just remove the SUM and you'll get it per row.

Answer 3

In addition to what @dnagirl answered, I would consider doing an extra step... adding a column to the table for "anyNulls". And set it as a flag for ANY empty/null values, so you would be able to quickly query those that need to be completed for data requirements instead of just how many of each category. One record could have all fields missing vs one column in multiple rows that need to be fixed.

Answer 4

As I understand it you have a table such as

id | FName | LName  | Age
1  | John  | ""     | NULL
2  | Mary  | Simons | NULL

And you want: row 1 has 2 empty/null fields, and row 2 has 1 empty/null fields

SELECT
    id,

    IF (FName IS NULL OR FName = '', 1, 0) + 
    IF (LName IS NULL OR LName = '', 1, 0) + 
    IF (Age IS NULL OR Age = '', 1, 0)
    as empty_field_count

Though now that I've written this I see Saul's response, I think that may have better performance that this solution.