Infragistics Forums

Hi,

I´m working on a spreadsheet like application where users can create its formulas just by typing it in a formula bar, according to our sintax and application data context. Recently, a customer has reported an error (as he sad) in the following formula:

SUM(A2:A10) , where values for A2 to A10 are: 3287.87, 502.72, 354.83, 857.21, 266.52, 33.32, 83.29, 43.37 and -5429.13

In this case, we just replace the text "A2:A10" by its values, resulting in "SUM(3287,87;502,72;354,83;857,21;266,52;33,32;83,29;43,37;-5429,13)" expression and then call the UltraCalcManager.Calculate method getting the unexpected value -9,09494701772928E-13.

I´m aware that this is a double data type imprecision issue but unfortunately I can´t use this as an argumente with my customer. And to make it worst, when he does the same in the MS Excel, he gets the expected result for formula SUM(A2:A10). It seems they do some kind of work around to avoid the inprecise result value.

Any suggestion?

 Att, Clodoaldo.

What result are you getting from Excel that is different from the CalcManager resut? When I try this formula, I get exactly the same result in both Excel and CalcManager:

By default, Excel will format the number differently and hide some of the decimal places, but if you apply a number format to the cell so that it shows all of the digits, it returns exactly the same result as the CalcManager does.