sorry I don't know the name for this problem, nor if it's even solvable.
I'm trying to calculate the new delivery cost based on previous data.
Say we have 2 products with respective delivery cost per item:
Apple --> 7 $ (7 -> current value)
Melon --> 20 $ (20 -> current value)
Then, I make a purchase of:
10 x Apple
50 x Melon
Total new delivery fee is: 1,130 $.
Just with the data above, is it possible to calculate and know that the new delivery fees are:
Apple --> 8 $ (10 x 8 $ = 80 $)
Melon --> 21 $ (50 x 21 $ = 1,050 $)
If the quantities were equal, then I could assume that 25.93% (7 / (7 + 20)) of the delivery fee belongs to Apples, and 74.07% belongs to Melons.
But since quantities are not equal, I have this problem of 2 percentages. I can't say that 25.93 % (based on previous delivery fee) or 16.67 % (based on quantity) of the total belongs to Apples.
This problem comes from making purchases of big and small items. If I just take the total and divide with all quantity, then the smaller items would have their delivery cost higher than their own cost.
Thanks in advance!
update - edit ------------
So I got this idea, it doesn't solve what I asked, but may be good enough for you as it is for me.
Basically, calculate the total delivery fee using old values:
10 x Apple ( 7 $) = 70 $
50 x Melon (20 $) = 1,000 $
total = 1,070 $
Then compare it with the new delivery fee: 1,130 $
1,070 -> 1,130 = 5.6 % Increase
Now I can use that % to update the products delivery fee:
Apple 7 $ + 5.6 % => 7.39 $
Melon 20 $ + 5.6 % => 21.12 $
This is not a solution for what I originally asked in the post (I asked poorly). But I think this might be what you were looking for if you had the same problem as I did.
As Lulu notes, you got the arithmetic in your example wrong.
In general, it's not possible, no. Here's how I know. There are two different per-apple and per-melon prices that produce the same total. So you cannot tell which of these (or some other) the right value is. For instance, in your example, the data
gives exactly the same total cost.
If you happen to have data for two different deliveries, say
then you can, generally, figure out the cost per apple or per melon. If you call the first of these $A$ and the second $M$, then we have
\begin{align} 10A + 50 M = 1070 \\ 8A + 32 M = 696 \end{align} We multiply the top equation by $8$ (the thing next to $A$ in the bottom one) and the bottom by 10 (the thing next to $A$ in the top one) to get \begin{align} 80A + 400 M = 8560 \\ 80A + 320 M = 6960 \end{align} We then subtract one equation from the other to get \begin{align} 80A + 400 M = 8560 \\ 0A + 80 M = 1600 \end{align} and now we can see that since 80M = 1600, we have $M = 20$. The first equation then becomes \begin{align} 80A + 400 \cdot 20 = 8560 \\ 80A + 8000= 8560 \\ 80A = 560 \\ 8A = 56 \\ A = 7 \\ \end{align}
This technique can fail: if one of the orders contains zero apples, it won't work (you'll need to do the "cross multiplying" using the numbers next to the $M$s instead!), and if the two orders are just multiples of one another, i.e. the first is $$ (4A, 3M) $$ and the second is $$ (8A, 6M) $$ then you won't have enough information. Otherwise, this technique should work (given two orders) to discover the hidden pricing.
Warning: Often, there a delivery charge, and that messes up everything!