Hi,
I am currently working on POC for migrating Oracle to SQL Server 2008 R2. Stuck with an issue related to rounding off calculations.
There is this "rate" value we are calculating using POWER functions (it has nested POWER functions actually). I am getting mismatches because the values are not correctly round off during calculation. There are around 258 rows where the values are 0.01 less than the required value (ex. I get the rate value as 6.31 where actually it should be 6.32).
In order to correct this issue, I changed the data types of the fields in the calculation from FLOAT to NUMERIC to correct out the precision. This is able to resolve the issue with the above 258 rows, but now I have another 61 rows where the value is 0.01 more than the required value (I get the rate value as 7.42 where actually it should be 7.41).
All the fields in the Oracle calculation is using NUMBER datatype (without any precision), so cant really figure out what equivalent data type should be used in SQL Server. Currently I am using FLOAT datatype.
Please advise if you have any pointers.
Regards
Manoj