Average Error: 0.0 → 0.0
Time: 1.3s
Precision: binary64
\[\mathsf{fma}\left(x - a, y - b, a \cdot b\right)\]
↓
\[\mathsf{fma}\left(x - a, y - b, a \cdot b\right)\]
\mathsf{fma}\left(x - a, y - b, a \cdot b\right)
↓
\mathsf{fma}\left(x - a, y - b, a \cdot b\right)
double code(double x, double a, double y, double b) {
	return ((double) fma(((double) (x - a)), ((double) (y - b)), ((double) (a * b))));
}
↓
double code(double x, double a, double y, double b) {
	return ((double) fma(((double) (x - a)), ((double) (y - b)), ((double) (a * b))));
}

Error

Bits error versus x

Bits error versus a

Bits error versus y

Bits error versus b

Try it out

Your Program's Arguments

Results

Enter valid numbers for all inputs

Derivation

  1. Initial program 0.0

    \[\mathsf{fma}\left(x - a, y - b, a \cdot b\right)\]

  2. Final simplification0.0

    \[\leadsto \mathsf{fma}\left(x - a, y - b, a \cdot b\right)\]

Reproduce

herbie shell --seed 2020152 
(FPCore (x a y b)
  :name "(fma (- x a) (- y b) (* a b))"
  :precision binary64
  (fma (- x a) (- y b) (* a b)))