Group by with having clause a key not equal - mysql

Question

I have a table structure and values as:-
used_items
id item_name qty
24 Potatoes 0
25 Potatoes 500
26 test 88
27 Wheat 3
28 abc 10
I want to group according to item_name such as query should return single item_name where qty!=0.
I have used this query-
SELECT * from used_items group by item_name having qty!='0'
This query doesn't return Potatoes, the result-
id item_name qty
26 test 88
27 Wheat 3
28 abc 10
Expected Result-
id item_name qty
25 Potatoes 500
26 test 88
27 Wheat 3
28 abc 10
It should also return the item_name Potatoes with non zero value including in the group.
How can I achieve this?
Edits-
I have used Group by because i want unique item_name and it should skip the item_name if qty is 0 for all same item_name. But the query skips if atleast one of the row qty is 0(such as Potatoes).
Also, if qty is 0, it should get the row which matches the item_name according to the id ascending. Hope i am clear, for eg;
The data:-
id item_name qty
24 Potatoes 0
25 Potatoes 500
27 Potatoes 400
28 test 88
29 Wheat 3
30 abc 10
Expected result, returning unique item_name with first inserted non-zero value
id item_name qty
25 Potatoes 500
28 test 88
29 Wheat 3
30 abc 10

Answer 1

The structure of your query is the problem. When you GROUP BY by you are working with aggregating data and the HAVING clause functions on an aggregate. Your query doesn't do any aggregation at the moment.
So you'd need to aggregate the qty by SUM or AVG or MAX in order to have an effective group by. I also wouldn't think about excluding the id here, because it's not useful in the aggregate.
For example:
SELECT item_name, MIN(qty)
FROM used_items
GROUP BY item_name
HAVING MIN(qty) <> '0'
You can also consider using DISTINCT if you just want to find unique combinations. Again, the id won't be useful with DISTINCT. DISTINCT does not aggregate.
For example:
SELECT DISTINCT item_name, qty
FROM used_items
WHERE qty <> '0';
Having a second look, if all you want to do is exclude all 0 qty rows, all you need is:
SELECT id, item_name, qty
WHERE qty <> '0';
Finally, just check that the data type of qty is actually a VARCHAR or CHAR like type. It really would make a lot more sense if it was an INT.
EDIT:
You can use a WHERE clause with a GROUP BY to exclude qty values that are 0. And since qty is an INT data type, you don't need to put values in quotes '.
Example:
SELECT item_name, MAX(qty)
FROM used_items
WHERE qty <> 0
GROUP BY item_name
I hope this helps.

Answer 2

If you want for each item_name the 1st non 0 (for the quantity) row then you can do it with a correlated subquery:
SELECT t.*
FROM tablename t
WHERE t.id = (SELECT MIN(id) FROM tablename WHERE item_name = t.item_name AND qty <> 0)
See the demo.
Results:
> id | item_name | qty
> -: | :-------- | --:
> 25 | Potatoes | 500
> 28 | test | 88
> 29 | Wheat | 3
> 30 | abc | 10

Answer 3

Why are you using the group by without an aggregate function? If I understood your question, to achieve your expected result you can just..
SELECT DISTINCT(item_name), qty from used_items where qty!='0

Answer 4

SELECT MAX(id) id, item_name, SUM(qty) qty
FROM used_items
GROUP BY item_name
HAVING SUM(qty) > 0
ORDER BY MAX(id);
should do it for you. See this. https://www.db-fiddle.com/f/9hmTPkNkhpjqGrFvUeLzjW/1
MAX(id)'s purpose is to get a valid id column value even from your GROUP BY operation.
You were misusing MySQL's notorious non-standard GROUP BY handling and therefore getting strange results.