The threshold for two measures to be considered equivalent if converted to the same unit. This is only needed due to floating-point error.
Returns the measure with the absolute value closest to positive infinity.
Returns the measure with the absolute value closest to negative infinity.