Hi folks.

I have a table that consists of 428 invoice records which have a total value of £477631.78. Now, I have been asked to try and identify a list of those invoices that amount to a total of £120796.27.

Is there any code/program out there which can be used to help identify all possible combinations of Invoice records that would total £120796.27?

Thanks in advance,
Kev

yes there are several programs out there which could do this, however none of them is going to do exactly what tyour want out of the box

SQL is a means of querying the data
however you may need to use another language as well to go through the permutations

you mention that you have to find £120K of invoice out of £480k

is there a reason for the 'magic' number, what is driving the request,
is it a series of invoices from a company / companies

Thanks for the swift response.

We kind of screwed up this weekend in that we didnt run a weekly update program that we were suppposed to. This has resulted in £120796.27 worth of Invoices being allocated a date of 3rd April 2006 instead of 1st April 2006.

Subsequently, this screws up our datawharehousing system, which is why we are wanting to move the correct Invoices back to a date of 1st April 2006.

Identifying those Invoice records that need to be moved back is proving to be a bit of a nightmare though.

Let's just imagine that you are a really lucky guy, and the £120796.27 discrepancy is composed of just 4 out of the 428 total invoices. That means you only have to check 33 billion different permutations.

Now imagine that you are not so lucky...do you feel lucky?

Now imagine you had made a backup...did you make a backup?

so its an invoicing problem
is the invoice number sequential? or is there some other method of identyfying the invoice sequence
is there a date range that works
what happens if you group by date
do you have a timestamp on the data
you jsut be 'lucky' and you data is in the sequence it was written
try a simple reportadding the value of invocies above + this pone and see if it gets to your magic number

Good point. If the data is sequential, he does not need to test permutions. He just needs to find the breakpoints that bracket the correct total amount.