mySQL Query Count Distinct and Percentage - mysql

Question

I'm having trouble creating a query that will report on training completions by users that are grouped by district within a region. For one training course the report needs to show for each district
Number of users assigned course
Number of users completed course
Percentage of users completed course
The output report should look like this (without the periods).:
District.........Assigned.......Completed....%
Arkansas..........20..............15...............75%
Illinois...............80..............80...............100%
Iowa.................10...............8.................80%
Michigan..........30..............20................66%
Here's the SQL query I have tried using
Select mytable.district as District,
(Select Count(Distinct mytable.user_id)
From mytable
Where mytable.district = District AND
mytable.training_title = 'My Course') As 'Assigned',
(Select Count(Distinct mytable.user_id)
From mytable
Where mytable.training_status = "Completed" AND
mytable.district = District AND
mytable.training_title = 'My Course') as 'Completed',
Concat(Round(100 * (Select Count(Distinct mytable.user_id)
From mytable
Where mytable.training_status = "Completed" AND
mytable.district = District AND
mytable.training_title = 'My Course') / (Select Count(Distinct mytable.user_id)
From mytable
Where
mytable.district = District AND
mytable.training_title = 'My Course'),0 ),"%") as '%'
From mytable
Where mytable.region = 'Midwest'
Group by District
It doesn't work at all like this. However if I substitute one of the district values (such as Arkansas) in for 'District' in the WHERE clause I can get the correct values for that district. However i need to find all of the districts for each region and calculate the values for district. This is only one of the regions I need to create the query for.
Two key issues:
All of the data exists in one table in my database. I have consolidated and imported it from another database.
The data contains duplicates Therefore I must use Distinct to eliminate the duplicates from the results.
Suggestions? Thanks in advance for your assistance!!!

Answer 1

As far as I can see, you could just sum things up as you want them and GROUP BY district. The subquery will take care of the duplicate rows.
SELECT
district,
COUNT(*) Assigned,
COUNT(CASE WHEN training_status='Completed' THEN 1 END) Completed,
CONCAT(COUNT(CASE WHEN training_status='Completed' THEN 1 END) /
COUNT(*) * 100, '%') `%`
FROM (
SELECT DISTINCT * FROM mytable
WHERE mytable.region = 'Midwest'
AND mytable.training_title = 'My Course') z
GROUP BY district;
An SQLfiddle to test with.