Average Error: 0.0 → 0.0
Time: 412.0ms
Precision: binary64
\[\mathsf{max}\left(\tanh x, 0.0\right)\]
↓
\[\mathsf{max}\left(\tanh x, 0.0\right)\]
\mathsf{max}\left(\tanh x, 0.0\right)
↓
\mathsf{max}\left(\tanh x, 0.0\right)
double code(double x) {
	return ((double) fmax(((double) tanh(x)), 0.0));
}
↓
double code(double x) {
	return ((double) fmax(((double) tanh(x)), 0.0));
}

Error

Bits error versus x

Try it out

Your Program's Arguments

Results

Enter valid numbers for all inputs

Derivation

  1. Initial program 0.0

    \[\mathsf{max}\left(\tanh x, 0.0\right)\]

  2. Final simplification0.0

    \[\leadsto \mathsf{max}\left(\tanh x, 0.0\right)\]

Reproduce

herbie shell --seed 2020153 
(FPCore (x)
  :name "(fmax (tanh x) 0.0)"
  :precision binary64
  (fmax (tanh x) 0.0))