Average Error: 3.1 → 3.1
Time: 862.0ms
Precision: binary64
\[\left\lfloorx \cdot \left(1 - \frac{b}{m}\right)\right\rfloor\]
↓
\[\left\lfloorx \cdot \left(1 - \frac{b}{m}\right)\right\rfloor\]
\left\lfloorx \cdot \left(1 - \frac{b}{m}\right)\right\rfloor
↓
\left\lfloorx \cdot \left(1 - \frac{b}{m}\right)\right\rfloor
double code(double x, double b, double m) {
	return ((double) floor(((double) (x * ((double) (1.0 - ((double) (b / m))))))));
}
↓
double code(double x, double b, double m) {
	return ((double) floor(((double) (x * ((double) (1.0 - ((double) (b / m))))))));
}

Error

Bits error versus x

Bits error versus b

Bits error versus m

Try it out

Your Program's Arguments

Results

Enter valid numbers for all inputs

Derivation

  1. Initial program 3.1

    \[\left\lfloorx \cdot \left(1 - \frac{b}{m}\right)\right\rfloor\]

  2. Final simplification3.1

    \[\leadsto \left\lfloorx \cdot \left(1 - \frac{b}{m}\right)\right\rfloor\]

Reproduce

herbie shell --seed 2020153 
(FPCore (x b m)
  :name "(floor (* x (- 1 (/ b m))))"
  :precision binary64
  (floor (* x (- 1.0 (/ b m)))))