Function round_decimal_f64 

fn round_decimal_f64(v: f64) -> f64

Rounds an f64 to [DECIMAL_PLACES] decimal places using pure arithmetic.

Eliminates the string formatting roundtrip entirely. For values like 4.6 that can’t be exactly represented in IEEE 754, this cleans up trailing noise (e.g. 4.6000000000000005 → 4.6).

Splits into integer and fractional parts before scaling to avoid overflow: large SAS datetime values (~1.9e9) multiplied by 1e14 would exceed f64’s exact integer range (2^53), causing rounding errors.