Hello labmaster,
the standard way of rounding will help you: multiply by 1000 (=3 digits after separator), round to nearest, divide by 1000.
But be aware: using floating point numbers can give "unwanted" results
as not all fractions can be represented by them! Example: comparision
of floating point numbers can give unexpected results...
Other solution:
If applicable you can use Integer numbers and treat them as fixed point
numbers. The last three digits represent the fractional part, all upper
digits the integer part.
Example: 12345 will be interpreted as 12.345, 9.876 will be represented
by 9876. But you need your own math routines (add/sub will work anyway,
mult/div needs correctional factors and so on).
Best regards,
GerdW

using LV2016/2019/2021 on Win10/11+cRIO, TestStand2016/2019