I agree this looks like a rounding issue. Update the formatting (on the column itself, and/or in the visualization) to display 15 decimal places of precision to verify if the data is indeed the same or to see if the data is in fact different (more likely). If the data is indeed different, then that is the explanation.

If the data does not appear different in the client, then consider the following 2 scenarios:

1) Spotfire only displays 15 decimal places of accuracy, so technically if the data had accruacy of more than 15 decimal places on import, it could be treated as different, despite appearing the same in the client. In this case, you can either modify your source data to match Spotfire's precision, truncate the data on import, or use Round() function to truncate data to match displayed value (like Round([mydata],15))

2) The application performs calculations using floating point math which inherently leads to some rounding errors (related search). There are some special cases where this means that some calculated data that 'appears' the same in Spotfire (even if the source data does not inidicate the difference) is treated as different/unique. The only workaround is to explicitly use the Round() function.

If either of those latter 2 scenarios are what you are seeing then you should open up a case with TIBCO Spotfire Support.

## 1 Comment

This looks like a rounding thing. Can you increase the decimal places on your percentage column? Perhaps some values are 7.781 and some are 7.782?