Reputation: 75
I have a data like this :
Table LOT
+-------+--------+
|Lot_id | Prod_id|
+-------+--------+
| LOT-1 | Prd-1 |
| LOT-1 | Prd-2 |
| LOT-1 | Prd-3 |
| LOT-2 | Prd-4 |
+-------+--------+
Table Process
+-------+--------+--------+------------+----------+
|proc_id|proc_cat|proc_seq|proc_prod_id|t_proc_qty|
+-------+--------+--------+------------+----------+
| 1 | Proc-A | 1 | Prd-1 | 100 |
| 2 | Proc-H | 2 | Prd-1 | 100 |
| 3 | Proc-D | 3 | Prd-1 | 100 |
| 4 | Proc-A | 1 | Prd-2 | 100 |
| 5 | Proc-H | 2 | Prd-2 | 100 |
| 6 | Proc-D | 3 | Prd-2 | 20 |
| 7 | Proc-Q | 4 | Prd-2 | 20 |
| 8 | Proc-A | 1 | Prd-3 | 100 |
| 9 | Proc-H | 2 | Prd-3 | 100 |
| 10 | Proc-D | 3 | Prd-3 | 50 |
| 11 | Proc-O | 1 | Prd-4 | 80 |
| 12 | Proc-F | 2 | Prd-4 | 80 |
| 13 | Proc-H | 3 | Prd-4 | 80 |
+-------+--------+--------+------------+----------+
And i want data like this if i want select just LOT=LOT-1. table LOT joined to table Process and data is accumulated sum(t_proc_qty) from last proc_seq each proc_prod_id and group by proc_cat and order by proc_seq
+--------+--------+----------+
|proc_cat|proc_seq|t_proc_qty|
+--------+--------+----------+
| Proc-D | 3 | 150 |->accumulated from Prd-1 and prd-3 in last process is seq 3
| Proc-Q | 4 | 20 |->accumulated from Prd-2 in last process is seq 4
+--------+--------+----------+
What queries I use in MySQL ?
I stucked in query
SELECT proc_cat, proc_seq, SUM(t_proc_qty)
FROM Process
LEFT JOIN Lot ON proc_prod_id=Prod_id
WHERE Lot_id='LOT-1'
GROUP BY proc_prod_id
ORDER BY proc_seq DESC LIMIT 1
this schema for trial query SQLFiddle
Upvotes: 2
Views: 65
Reputation: 94894
From table Process
you want the records with the highest proc_id
per proc_prod_id
:
select *
from process
where not exists
(
select *
from process later
where later.proc_prod_id = process.proc_prod_id
and later.proc_id > process.proc_id
);
From this data you want an aggregate per proc_cat
and proc_sec
. And you also want to consider only prod_id
for 'LOT-1' in table LOT
.
The complete query:
select proc_cat, proc_seq, sum(t_proc_qty)
from process
where proc_prod_id in (select prod_id from lot where lot_id = 'LOT-1')
and not exists
(
select *
from process later
where later.proc_prod_id = process.proc_prod_id
and later.proc_id > process.proc_id
)
group by proc_cat, proc_seq
order by proc_cat, proc_seq;
SQL fiddle: http://sqlfiddle.com/#!9/1fa3fd/5
Upvotes: 1