I would like to highlight recent issues I have been having with Aspose handling of the CEILING function with significance parameter of 0.12.
A client supplied a hugely complex spreadsheet which produced incorrect results when run through Aspose Cells when compared to running directly in Excel. The spreadsheet is client confidential so cannot be supplied.
I would be grateful to receive any feedback on these observations., thank you
The workbook had many worksheets including 720 formula including CEILING function with significance parameter of 0.12,
i.e. =CEILING(F90-IF(Inputs!F$192=“No”,0,‘Pen Deb PRE DOL LR’!F322),0.12)
Results with CEILING(xxx,0.12)
9371.16,10664.52,645.12,4514.88,0,0,0,0,NULL,0,0,0,0,125000
I modified the SS to use CEILING(xxx,0.01) everywhere as an alternative. Results:
8835.48,9986.03,603.08,4221.5,535.470000000001,678.220000000001,41.89,293.27,NULL,526.67,684.12,42.2900000000001,296.08,125000
I also modified the SS to use MROUND(0.05999 + xxx,0.12) which should be mathematically more or less equivalent. Results:
8835.6,9986.28,603.12,4221.6,535.439999999999,678.24,41.88,293.280000000001,NULL,526.68,684.12,42.24,296.04,125000
As you can see the results with CEILING(xxx,0.12) are considerably different to those using MROUND(0.05999 + xxx,0.12) , which are very close to the results obtained when running the original spreadsheet directly in Excel.
(This is tested using Aspose Cells 22.5)