MySQL value near by, join, abs with dynamic value - mysql

Question

The goal is the get the price of the nearest number of the m3 and weight and put this in a table.
This sample works for me. But i hard code m3,weight,country and transportername
Query1:
SELECT
country,
transportername,
ABS(m3-0.5) as ABS_M3 ,
m3,
ABS(weight-5) as ABS_weight,
weight,
price
FROM database.transportcost
WHERE m3 >= 0.5 AND
weight >= 5 AND
country = "GB" AND
transportername like '%aa%'
Group by transportername
Order by ABS(m3-0.5) +
ABS(weight-5) +
price
Now i want to get the hard code value from a other table. I don't know what value i need to put in ~~xx.volumen~~ and ~~xx.gewicht~~.
Query2:
INSERT INTO database.transportcostPerItem (sku,country,transportname,weight,m3,price)
SELECT g.orderitemid, p2.country, p2.transportername, p2.weight, p2.m3, p2.price
FROM database.orderitem g
JOIN database.order i on i.orderid = g.orderid
JOIN database.matrixtable xx on g.orderitemid = xx.sku
JOIN database.vlog cc on i.orderid = cc.orderid
JOIN
(
SELECT *
FROM database.transportcost
Group by transportername
Order by ABS(m3-~~xx.volumen~~) +
ABS(weight-~~xx.gewicht~~) +
price
) as p2 on p2.country = i._country AND
p2.transportername = cc._transportername AND
p2.weight = xx.gewicht AND
p2.m3 = xx.volumen
;
Edit:
Sample data
Table: database.transportcost
country | transportername | m3 | weight | price
--------+-----------------+--------+--------+--------
GB | aa | 50.00 | 50 | 77.70
GB | bb | 0.50 | 125 | 83.19
GB | cc | 0.50 | 125 | 96.03
GB | bb | 0.60 | 150 | 83.19
GB | aa | 75.00 | 75 | 89.04
GB | cc | 0.60 | 150 | 96.03
GB | dd | 50.00 | 50 | 163.38
GB | cc | 0.70 | 175 | 96.03
GB | bb | 0.70 | 175 | 96.85
GB | ee | 0.53 | 175 | 102.78
GB | gg | 0.53 | 175 | 110.37
GB | aa | 100.00 | 100 | 89.04
GB | cc | 0.80 | 200 | 96.03
GB | bb | 0.80 | 200 | 96.85
GB | ff | 0.60 | 200 | 110.33
DE | aaa | 0.50 | 50 | 26.40
DE | bbb | 0.50 | 75 | 31.84
DE | aaa | 0.75 | 75 | 34.19
DE | ccc | 0.50 | 100 | 34.17
DE | bbb | 0.67 | 100 | 35.47
DE | ccc | 1.00 | 100 | 37.59
DE | ddd | 0.50 | 100 | 62.38
DE | ccc | 0.63 | 125 | 43.04
DE | bbb | 0.83 | 125 | 44.24
DE | aaa | 0.93 | 125 | 45.84
DE | eee | 0.50 | 125 | 53.80
DE | fff | 0.50 | 125 | 54.02
Query 1 results for the country GB
country | transportername | ABS_M3 | m3 | ABS_weight | weight | price
--------+-----------------+-----------------------+-------+------------+--------+-------
GB | aa | 49.5 | 50.00 | 45 | 50 | 77.70
GB | bb | 0 | 0.50 | 120 | 125 | 83.19
GB | cc | 0 | 0.50 | 120 | 125 | 96.03
GB | dd | 49.5 | 50.00 | 45 | 50 | 163.38
GB | ee | 0.030000000000000027 | 0.53 | 170 | 175 | 102.78
GB | gg | 0.030000000000000027 | 0.53 | 170 | 175 | 110.37
Query 1 results for the country DE
country | transportername | ABS_M3 | m3 | ABS_weight | weight | price
--------+-----------------+--------+------+------------+--------+-------
DE | aaa | 0 | 0.50 | 45 | 50 | 26.40
DE | bbb | 0 | 0.50 | 70 | 75 | 31.84
DE | ccc | 0 | 0.50 | 95 | 100 | 34.17
DE | ddd | 0 | 0.50 | 95 | 100 | 62.38
DE | eee | 0 | 0.50 | 120 | 125 | 53.80
DE | fff | 0 | 0.50 | 120 | 125 | 54.02
DE | ggg | 0 | 0.50 | 195 | 200 | 87.29
Results in the end Query 2 should look like this:
Table: database.transportcostPerItem
orderitemid | country | transportername | m3 | weight | price
------------+---------+-----------------+-----+--------+------
1 | GB | aa | 0,5 | 5 | 77.70
2 | DE | aaa | 0,5 | 5 | 26.40

Answer 1

It seems you are merely looking for the cheapest transport for given m3 and weight. In a first step you'd find all transports for at least the m3 and weight given. Of these you pick the lowest price then.
select *
from transportcost
where country = 'GB'
and m3 >= 0.5
and weight >= 5
order by price
limit 1;
If you wanted the measures closest to your given m3 and weight instead, even if that transport would be more expensive, you could order by the sum of the percentages:
select *
from transportcost
where country = 'GB'
and m3 >= 0.5
and weight >= 5
order by m3 / 0.5 + weight / 5, price
limit 1;
Now instead of given m3, weight, country, and transporter name, you get the values from an order details table. If this were just about a single order details record, the approach would be exactly the same, only instead of comparing with fixed values (m3 >= 0.5) you'd compare with the order details record's values (transportcost.m3 >= orderdetail.m3). Alas, with multiple order details records this approach doesn't work anymore, because we cannot limit the results to one row, but need one row per order details record instead. This would be solved with window functions (ROW_NUMBER, RANK etc.) or lateral joins (CROSS APPLY) in standard SQL. MySQL features neither.
Here is how to join the transport cost table:
select *
from <your order tables>
join transportcost tc
where tc.country = i._country
and tc._transportername = cc._transportername
and tc.m3 >= xx.volumen
and tc.weight >= xx.gewicht;
And now you must find a way to rank your results, so as to only get the best match per order detail. One way is to emulate ROW_NUMBER with variables. You may want to look this up in other answers. Another would be a limit subquery in the SELECT clause, but for this to work you need a single column identifying a record in the transport cost table. Let's say you add a column called ID and fill it with unique values...
select <some order columns>, tc.*
from
(
select <some order columns>,
(
select id
from transportcost tc
where tc.country = i._country
and tc._transportername = cc._transportername
and tc.m3 >= xx.volumen
and tc.weight >= xx.gewicht
order by tc.m3 / xx.volumen + tc.weight / xx.gewicht, tc.price
limit 1
) as best_transportcost_id
from <your order tables>
) data
join transportcost tc
where tc.id = data.best_transportcost_id;
(You could do this without an ID somehow by concatenating the values, e.g. 'GB-aa-50.00-50-77.70', so as to have a string identifying a transport cost record, but I don't recommend this. It's better to have a unique ID with an index for quick lookups. And even better would be to work with a better DBMS :-)

Answer 2

I used the solution of #Thorsten.
This is the results
INSERT INTO database.transportcostperitem (transportcostID, orderID, sku, country, transportername, weight, m3,price,shippingmethod,shippingstatus,shippingDate)
SELECT
transportcostID,
orderID,
sku,
_country,
transportername,
transportTmp._weightReal,
transportTmp._m3 ,
price,
_shippingMethod,
_shippingStatus,
_shippingDate
FROM(
SELECT
b.orderID,
o.sku,
b.transporter,
o._country,
i._shippingMethod,
i._shippingStatus,
i._shippingDate,
o._m3,
o._weightReal,
(
SELECT id
FROM database.transportcost tc
where
tc.transportername = b._transporterName AND
tc.country = o._country AND
tc.weight >= o._weightReal AND
tc.m3 >= o._m3
order by o._m3 / tc.m3 + o._weightReal / tc.weight , tc.price
limit 1
) as transportcostID
FROM database.orderitem o
JOIN database.order i on i.orderID = o.orderID
JOIN database.vlog b on b.orderID = o.orderID
WHERE sku >0
)as transportTmp
JOIN database.transportcost tc
where tc.id = transportcostID AND
_shippingMethod is not null
orderID by sku
;