Maksimov and Kolovsky, Equation (4)

Average Error: 26.89% → 0.13%

Time: 13.4s

Precision: binary64

Cost: 13504

Math

\[\left(J \cdot \left(e^{\ell} - e^{-\ell}\right)\right) \cdot \cos \left(\frac{K}{2}\right) + U \]

↓

\[\left(2 \cdot \sinh \ell\right) \cdot \left(J \cdot \cos \left(K \cdot 0.5\right)\right) + U \]

FPCore

(FPCore (J l K U)
 :precision binary64
 (+ (* (* J (- (exp l) (exp (- l)))) (cos (/ K 2.0))) U))

↓

(FPCore (J l K U)
 :precision binary64
 (+ (* (* 2.0 (sinh l)) (* J (cos (* K 0.5)))) U))

C

double code(double J, double l, double K, double U) {
	return ((J * (exp(l) - exp(-l))) * cos((K / 2.0))) + U;
}

↓

double code(double J, double l, double K, double U) {
	return ((2.0 * sinh(l)) * (J * cos((K * 0.5)))) + U;
}

Fortran

real(8) function code(j, l, k, u)
    real(8), intent (in) :: j
    real(8), intent (in) :: l
    real(8), intent (in) :: k
    real(8), intent (in) :: u
    code = ((j * (exp(l) - exp(-l))) * cos((k / 2.0d0))) + u
end function

↓

real(8) function code(j, l, k, u)
    real(8), intent (in) :: j
    real(8), intent (in) :: l
    real(8), intent (in) :: k
    real(8), intent (in) :: u
    code = ((2.0d0 * sinh(l)) * (j * cos((k * 0.5d0)))) + u
end function

Java

public static double code(double J, double l, double K, double U) {
	return ((J * (Math.exp(l) - Math.exp(-l))) * Math.cos((K / 2.0))) + U;
}

↓

public static double code(double J, double l, double K, double U) {
	return ((2.0 * Math.sinh(l)) * (J * Math.cos((K * 0.5)))) + U;
}

Python

def code(J, l, K, U):
	return ((J * (math.exp(l) - math.exp(-l))) * math.cos((K / 2.0))) + U

↓

def code(J, l, K, U):
	return ((2.0 * math.sinh(l)) * (J * math.cos((K * 0.5)))) + U

Julia

function code(J, l, K, U)
	return Float64(Float64(Float64(J * Float64(exp(l) - exp(Float64(-l)))) * cos(Float64(K / 2.0))) + U)
end

↓

function code(J, l, K, U)
	return Float64(Float64(Float64(2.0 * sinh(l)) * Float64(J * cos(Float64(K * 0.5)))) + U)
end

MATLAB

function tmp = code(J, l, K, U)
	tmp = ((J * (exp(l) - exp(-l))) * cos((K / 2.0))) + U;
end

↓

function tmp = code(J, l, K, U)
	tmp = ((2.0 * sinh(l)) * (J * cos((K * 0.5)))) + U;
end

Wolfram

code[J_, l_, K_, U_] := N[(N[(N[(J * N[(N[Exp[l], $MachinePrecision] - N[Exp[(-l)], $MachinePrecision]), $MachinePrecision]), $MachinePrecision] * N[Cos[N[(K / 2.0), $MachinePrecision]], $MachinePrecision]), $MachinePrecision] + U), $MachinePrecision]

↓

code[J_, l_, K_, U_] := N[(N[(N[(2.0 * N[Sinh[l], $MachinePrecision]), $MachinePrecision] * N[(J * N[Cos[N[(K * 0.5), $MachinePrecision]], $MachinePrecision]), $MachinePrecision]), $MachinePrecision] + U), $MachinePrecision]

Derivation

1. Initial program 26.89

   \[\left(J \cdot \left(e^{\ell} - e^{-\ell}\right)\right) \cdot \cos \left(\frac{K}{2}\right) + U \]

2. Simplified 26.89

   \[\leadsto \color{blue}{\mathsf{fma}\left(e^{\ell} - e^{-\ell}, J \cdot \cos \left(\frac{K}{2}\right), U\right)} \]
   Proof

   [Start] 26.89	\[ \left(J \cdot \left(e^{\ell} - e^{-\ell}\right)\right) \cdot \cos \left(\frac{K}{2}\right) + U \]
   *-commutative [=>] 26.89	\[ \color{blue}{\cos \left(\frac{K}{2}\right) \cdot \left(J \cdot \left(e^{\ell} - e^{-\ell}\right)\right)} + U \]
   associate-*r* [=>] 26.89	\[ \color{blue}{\left(\cos \left(\frac{K}{2}\right) \cdot J\right) \cdot \left(e^{\ell} - e^{-\ell}\right)} + U \]
   *-commutative [<=] 26.89	\[ \color{blue}{\left(e^{\ell} - e^{-\ell}\right) \cdot \left(\cos \left(\frac{K}{2}\right) \cdot J\right)} + U \]
   fma-def [=>] 26.89	\[ \color{blue}{\mathsf{fma}\left(e^{\ell} - e^{-\ell}, \cos \left(\frac{K}{2}\right) \cdot J, U\right)} \]
   *-commutative [=>] 26.89	\[ \mathsf{fma}\left(e^{\ell} - e^{-\ell}, \color{blue}{J \cdot \cos \left(\frac{K}{2}\right)}, U\right) \]

3. Applied egg-rr 0.13

   \[\leadsto \color{blue}{\left(2 \cdot \sinh \ell\right) \cdot \left(J \cdot \cos \left(K \cdot 0.5\right)\right) + U} \]

4. Final simplification 0.13

   \[\leadsto \left(2 \cdot \sinh \ell\right) \cdot \left(J \cdot \cos \left(K \cdot 0.5\right)\right) + U \]

Alternatives

Alternative 1
Error	0.8%
Cost	7488

\[U + \frac{-2}{\ell \cdot 0.16666666666666666 + \frac{-1}{\ell}} \cdot \left(J \cdot \cos \left(K \cdot 0.5\right)\right) \]

Alternative 2
Error	0.87%
Cost	7488

\[U + \frac{2 \cdot \left(J \cdot \cos \left(K \cdot 0.5\right)\right)}{\ell \cdot -0.16666666666666666 + \frac{1}{\ell}} \]

Alternative 3
Error	0.96%
Cost	7104

\[U + 2 \cdot \left(\ell \cdot \left(J \cdot \cos \left(K \cdot 0.5\right)\right)\right) \]

Alternative 4
Error	13.13%
Cost	6848

\[U + \left(2 \cdot \sinh \ell\right) \cdot J \]

Alternative 5
Error	13.54%
Cost	832

\[U + 2 \cdot \frac{J}{\ell \cdot -0.16666666666666666 + \frac{1}{\ell}} \]

Alternative 6
Error	29.36%
Cost	452

\[\begin{array}{l} \mathbf{if}\;J \leq -2.25 \cdot 10^{+133}:\\ \;\;\;\;\ell \cdot \left(2 \cdot J\right)\\ \mathbf{else}:\\ \;\;\;\;U\\ \end{array} \]

Alternative 7
Error	13.57%
Cost	448

\[U + J \cdot \left(2 \cdot \ell\right) \]

Alternative 8
Error	28.84%
Cost	64

\[U \]

Reproduce

herbie shell --seed 2023089 
(FPCore (J l K U)
  :name "Maksimov and Kolovsky, Equation (4)"
  :precision binary64
  (+ (* (* J (- (exp l) (exp (- l)))) (cos (/ K 2.0))) U))