DSUM Database "Subtraction"

[isadoko]

XL: 2003
OS: XP

I have a table several 1000 rows in length, headed thus:

Office No Date Amount Source B385# (1,4,5)
1 01/20/05 76,752.98 1
1 01/20/05 (34,291.57) 4
1 01/20/05 (6,304.15) 5

I have a criteria table that looks thus:

Office No Date Date Source B385# (1,4,5)
1 >=1/1/05 <=1/31/05 1
2 >=1/1/05 <=1/31/05 1
3 >=1/1/05 <=1/31/05 1

I want to use only one header in my criteria table and instead have the last or next record reference the previous one and "subtract" one record to keep track of which record to manipulate. I have tried this formula (and variations therof):

=DSUM('Ledger'!$A:$E,"Source B385# (1,4,5)",'DSUM'!A1:D3)-DSUM('Ledger'!$A:$E,"Source B385# (1,4,5)",'DSUM'!A1:D2)

...and it yielded the wrong answer. I want to forego this:

Office No Date Date Source B385# (1,4,5)
1 >=1/1/05 <=1/31/05 1
Office No Date Date Source B385# (1,4,5)
2 >=1/1/05 <=1/31/05 1
Office No Date Date Source B385# (1,4,5)
3 >=1/1/05 <=1/31/05 1

Would apprieciate some hint at a better, working solution. Thanks much.

 

[Aladin Akyurek]

See for various mehods for invoking d-functions:

http://www.mrexcel.com/board2/viewtopic.php?t=60895&postdays=0&postorder=asc&start=0

which also includes a link to MS regarding data tables.

 

[isadoko]

Aladin,

Thanks for the response: it was good learning. However, I found it unable to give me the solution I seek. Using the Microsoft How To solution (and changing to DCOUNTA) I quickly counted the number of instances of my B385 codes: 1, 4, 5 by my 78 offices; what I could not do is extract the data by date range. This I need for my data will run from Jan-Dec and I need it by month.

My original criteria was Office No, Date, Date, B385# (1,4,5), and it works fine except as you know I had to repeat it for all 78 offices for a total of 168 rows of criteria per month.

What I really desire is a smaller criteria table. One that takes account of the B385 codes, date range and offices; and drop from one row to the next automatically. I hope I am explaining myself clearly enough as I am not an Excel expert by any measure: a true novice willing to experiment and learn. Thanks again.