Hi Ravish,
You could try rounding or truncating your input data to a specific number of digits. My gut tells me that the engine is finding minor precision differences past what is being shown in column E. If you only are analyzing to the nth decimal place, then consider rounding or truncating to n+1 to help eliminate this finding.
Thanks,
Trevor
Product Manager
Hi Trevor,
Thank you for the reply.
Truncating inputs worked. When aval and base values are truncated to 6 decimal places the error seems to go away. However, the study we are working on requires the original values without truncation.
Can we assume this is a bug in the application and document it in ADRG?. Please give your suggestions.
Thanks,
Ravish
Ravish,
Yes, I would suggest you document this in the ADRG. The reason this occurs is that SAS uses a floating point arithmetic which is slightly different than the way our Java application does the calculation. The differences are minor and not visibly seen (as they occur far out past the decimal). This is why rounding/truncating to <7 digits will solve the issue on your end.
Thanks,
Trevor
Product Manager
Dear Ravish,
You should NEVER change any data in your input for the sake of passing validation. That is the world upside down.
It might be that you would like to truncate your data for other reasons, but surely not for compensating a problem of the validation software.
Hi there,
Pinnacle report for community version 2.2.0 & 3.0.0 is showing an error CHG!=AVAL-BASE even though the calculated value of CHG is correct. Please refer the screenshot.
Can you let me know if this is a known issue or is there any workaround for this?.
Thanks,
Ravish