Hi All,

Im looking for the best solution for this.

We have 10 years of data (Invoices), there is only 1 LineItem per Invoice
with a possibility of 1 of 36 Types of Items

There are 4 Groups of Items, > PT , S, L, F ( representing the type )
with 6 terms > 3, 6, 12, 24, 36, 48 ( representing a block of time in
months )
an ID ( our account number )
and a Invoice Date

So the Invoice file might look like this

TYPE DATE ID
3 S 1/23/2005 54875957
48 L 12/3/2000 32862547
24 PT 6/4/2001 24587596
etc.. ( around 300,000 records )

In the contacts I need to summarized the data (Per ID ) , (Per Year), (Per
Month) / then also (Per ID ) , (Per Year), ( Per Type )

Not all contacts will need this, since some wont have invoices dating years
back

Trying to keep this from having a TON of calcs that are summarizing every
time, since all prior years invoice # ( count ) will never change or need
updating

Im trying to figure out maybe a scripted way to calc these, maybe set a new
record in a seperate table, that upon running updates the record.

What Im trying to show, is the number of contracts per month sold per year,
so we can see if a there is more or less sold in specific months, with a
breakdown of types

Help, My brain is getting fried.

Thanks

Just a few remarks :
- scripts run faster than calc fields - sometimes I measured a 10 times fold
(FM6) -.
- export is probably the best approach as it is generally speaking pretty
fast. You even may set up a number of partial raw export-import to split the
problem into pieces, and thus save time.
- you may - or may not - have some use of the 'Get(Summary)' function, but
as it is based on a sort and is normally a calc, it must be used only on
subsets of your large number of records.
- I feel the key is to carefully choose the subsets that are 1) useful and
2) practical for handling, and then run them one after the other.
Just my 2 cents.
Remi-Noel


"EagleSA" <development (AT) Xeaglewarranty (DOT) com> a écrit dans le message de news:
5WDhh.9182$cx4.661 (AT) trndny09 (DOT) ..