HairBSDF, sample_f, cosTheta

Percentage Accurate: 99.5% → 99.5%

Time: 8.4s

Alternatives: 17

Speedup: 1.0×

Specification

\[\left(10^{-5} \leq u \land u \leq 1\right) \land \left(0 \leq v \land v \leq 109.746574\right)\]

Math

\[\begin{array}{l} \\ 1 + v \cdot \log \left(u + \left(1 - u\right) \cdot e^{\frac{-2}{v}}\right) \end{array} \]

FPCore

(FPCore (u v)
 :precision binary32
 (+ 1.0 (* v (log (+ u (* (- 1.0 u) (exp (/ -2.0 v))))))))

C

float code(float u, float v) {
	return 1.0f + (v * logf((u + ((1.0f - u) * expf((-2.0f / v))))));
}

Fortran

module fmin_fmax_functions
    implicit none
    private
    public fmax
    public fmin

    interface fmax
        module procedure fmax88
        module procedure fmax44
        module procedure fmax84
        module procedure fmax48
    end interface
    interface fmin
        module procedure fmin88
        module procedure fmin44
        module procedure fmin84
        module procedure fmin48
    end interface
contains
    real(8) function fmax88(x, y) result (res)
        real(8), intent (in) :: x
        real(8), intent (in) :: y
        res = merge(y, merge(x, max(x, y), y /= y), x /= x)
    end function
    real(4) function fmax44(x, y) result (res)
        real(4), intent (in) :: x
        real(4), intent (in) :: y
        res = merge(y, merge(x, max(x, y), y /= y), x /= x)
    end function
    real(8) function fmax84(x, y) result(res)
        real(8), intent (in) :: x
        real(4), intent (in) :: y
        res = merge(dble(y), merge(x, max(x, dble(y)), y /= y), x /= x)
    end function
    real(8) function fmax48(x, y) result(res)
        real(4), intent (in) :: x
        real(8), intent (in) :: y
        res = merge(y, merge(dble(x), max(dble(x), y), y /= y), x /= x)
    end function
    real(8) function fmin88(x, y) result (res)
        real(8), intent (in) :: x
        real(8), intent (in) :: y
        res = merge(y, merge(x, min(x, y), y /= y), x /= x)
    end function
    real(4) function fmin44(x, y) result (res)
        real(4), intent (in) :: x
        real(4), intent (in) :: y
        res = merge(y, merge(x, min(x, y), y /= y), x /= x)
    end function
    real(8) function fmin84(x, y) result(res)
        real(8), intent (in) :: x
        real(4), intent (in) :: y
        res = merge(dble(y), merge(x, min(x, dble(y)), y /= y), x /= x)
    end function
    real(8) function fmin48(x, y) result(res)
        real(4), intent (in) :: x
        real(8), intent (in) :: y
        res = merge(y, merge(dble(x), min(dble(x), y), y /= y), x /= x)
    end function
end module

real(4) function code(u, v)
use fmin_fmax_functions
    real(4), intent (in) :: u
    real(4), intent (in) :: v
    code = 1.0e0 + (v * log((u + ((1.0e0 - u) * exp(((-2.0e0) / v))))))
end function

Julia

function code(u, v)
	return Float32(Float32(1.0) + Float32(v * log(Float32(u + Float32(Float32(Float32(1.0) - u) * exp(Float32(Float32(-2.0) / v)))))))
end

MATLAB

function tmp = code(u, v)
	tmp = single(1.0) + (v * log((u + ((single(1.0) - u) * exp((single(-2.0) / v))))));
end

Initial Program: 99.5% accurate, 1.0× speedup

Math

\[\begin{array}{l} \\ 1 + v \cdot \log \left(u + \left(1 - u\right) \cdot e^{\frac{-2}{v}}\right) \end{array} \]

FPCore

(FPCore (u v)
 :precision binary32
 (+ 1.0 (* v (log (+ u (* (- 1.0 u) (exp (/ -2.0 v))))))))

C

float code(float u, float v) {
	return 1.0f + (v * logf((u + ((1.0f - u) * expf((-2.0f / v))))));
}

Fortran

module fmin_fmax_functions
    implicit none
    private
    public fmax
    public fmin

    interface fmax
        module procedure fmax88
        module procedure fmax44
        module procedure fmax84
        module procedure fmax48
    end interface
    interface fmin
        module procedure fmin88
        module procedure fmin44
        module procedure fmin84
        module procedure fmin48
    end interface
contains
    real(8) function fmax88(x, y) result (res)
        real(8), intent (in) :: x
        real(8), intent (in) :: y
        res = merge(y, merge(x, max(x, y), y /= y), x /= x)
    end function
    real(4) function fmax44(x, y) result (res)
        real(4), intent (in) :: x
        real(4), intent (in) :: y
        res = merge(y, merge(x, max(x, y), y /= y), x /= x)
    end function
    real(8) function fmax84(x, y) result(res)
        real(8), intent (in) :: x
        real(4), intent (in) :: y
        res = merge(dble(y), merge(x, max(x, dble(y)), y /= y), x /= x)
    end function
    real(8) function fmax48(x, y) result(res)
        real(4), intent (in) :: x
        real(8), intent (in) :: y
        res = merge(y, merge(dble(x), max(dble(x), y), y /= y), x /= x)
    end function
    real(8) function fmin88(x, y) result (res)
        real(8), intent (in) :: x
        real(8), intent (in) :: y
        res = merge(y, merge(x, min(x, y), y /= y), x /= x)
    end function
    real(4) function fmin44(x, y) result (res)
        real(4), intent (in) :: x
        real(4), intent (in) :: y
        res = merge(y, merge(x, min(x, y), y /= y), x /= x)
    end function
    real(8) function fmin84(x, y) result(res)
        real(8), intent (in) :: x
        real(4), intent (in) :: y
        res = merge(dble(y), merge(x, min(x, dble(y)), y /= y), x /= x)
    end function
    real(8) function fmin48(x, y) result(res)
        real(4), intent (in) :: x
        real(8), intent (in) :: y
        res = merge(y, merge(dble(x), min(dble(x), y), y /= y), x /= x)
    end function
end module

real(4) function code(u, v)
use fmin_fmax_functions
    real(4), intent (in) :: u
    real(4), intent (in) :: v
    code = 1.0e0 + (v * log((u + ((1.0e0 - u) * exp(((-2.0e0) / v))))))
end function

Julia

function code(u, v)
	return Float32(Float32(1.0) + Float32(v * log(Float32(u + Float32(Float32(Float32(1.0) - u) * exp(Float32(Float32(-2.0) / v)))))))
end

MATLAB

function tmp = code(u, v)
	tmp = single(1.0) + (v * log((u + ((single(1.0) - u) * exp((single(-2.0) / v))))));
end

Alternative 1: 99.5% accurate, 1.0× speedup

Math

\[\begin{array}{l} \\ 1 + v \cdot \log \left(u + \left(1 - u\right) \cdot e^{\frac{-2}{v}}\right) \end{array} \]

FPCore

(FPCore (u v)
 :precision binary32
 (+ 1.0 (* v (log (+ u (* (- 1.0 u) (exp (/ -2.0 v))))))))

C

float code(float u, float v) {
	return 1.0f + (v * logf((u + ((1.0f - u) * expf((-2.0f / v))))));
}

Fortran

module fmin_fmax_functions
    implicit none
    private
    public fmax
    public fmin

    interface fmax
        module procedure fmax88
        module procedure fmax44
        module procedure fmax84
        module procedure fmax48
    end interface
    interface fmin
        module procedure fmin88
        module procedure fmin44
        module procedure fmin84
        module procedure fmin48
    end interface
contains
    real(8) function fmax88(x, y) result (res)
        real(8), intent (in) :: x
        real(8), intent (in) :: y
        res = merge(y, merge(x, max(x, y), y /= y), x /= x)
    end function
    real(4) function fmax44(x, y) result (res)
        real(4), intent (in) :: x
        real(4), intent (in) :: y
        res = merge(y, merge(x, max(x, y), y /= y), x /= x)
    end function
    real(8) function fmax84(x, y) result(res)
        real(8), intent (in) :: x
        real(4), intent (in) :: y
        res = merge(dble(y), merge(x, max(x, dble(y)), y /= y), x /= x)
    end function
    real(8) function fmax48(x, y) result(res)
        real(4), intent (in) :: x
        real(8), intent (in) :: y
        res = merge(y, merge(dble(x), max(dble(x), y), y /= y), x /= x)
    end function
    real(8) function fmin88(x, y) result (res)
        real(8), intent (in) :: x
        real(8), intent (in) :: y
        res = merge(y, merge(x, min(x, y), y /= y), x /= x)
    end function
    real(4) function fmin44(x, y) result (res)
        real(4), intent (in) :: x
        real(4), intent (in) :: y
        res = merge(y, merge(x, min(x, y), y /= y), x /= x)
    end function
    real(8) function fmin84(x, y) result(res)
        real(8), intent (in) :: x
        real(4), intent (in) :: y
        res = merge(dble(y), merge(x, min(x, dble(y)), y /= y), x /= x)
    end function
    real(8) function fmin48(x, y) result(res)
        real(4), intent (in) :: x
        real(8), intent (in) :: y
        res = merge(y, merge(dble(x), min(dble(x), y), y /= y), x /= x)
    end function
end module

real(4) function code(u, v)
use fmin_fmax_functions
    real(4), intent (in) :: u
    real(4), intent (in) :: v
    code = 1.0e0 + (v * log((u + ((1.0e0 - u) * exp(((-2.0e0) / v))))))
end function

Julia

function code(u, v)
	return Float32(Float32(1.0) + Float32(v * log(Float32(u + Float32(Float32(Float32(1.0) - u) * exp(Float32(Float32(-2.0) / v)))))))
end

MATLAB

function tmp = code(u, v)
	tmp = single(1.0) + (v * log((u + ((single(1.0) - u) * exp((single(-2.0) / v))))));
end

Derivation

1. Initial program 99.6%

   \[1 + v \cdot \log \left(u + \left(1 - u\right) \cdot e^{\frac{-2}{v}}\right) \]
2. Add Preprocessing
3. Add Preprocessing

Alternative 2: 99.5% accurate, 1.0× speedup

Math

\[\begin{array}{l} \\ \mathsf{fma}\left(\log \left(\left(1 - u\right) \cdot e^{\frac{-2}{v}} + u\right), v, 1\right) \end{array} \]

FPCore

(FPCore (u v)
 :precision binary32
 (fma (log (+ (* (- 1.0 u) (exp (/ -2.0 v))) u)) v 1.0))

C

float code(float u, float v) {
	return fmaf(logf((((1.0f - u) * expf((-2.0f / v))) + u)), v, 1.0f);
}

Julia

function code(u, v)
	return fma(log(Float32(Float32(Float32(Float32(1.0) - u) * exp(Float32(Float32(-2.0) / v))) + u)), v, Float32(1.0))
end

Derivation

1. Initial program 99.6%

   \[1 + v \cdot \log \left(u + \left(1 - u\right) \cdot e^{\frac{-2}{v}}\right) \]
2. Add Preprocessing
3. Step-by-step derivation

   1. lift-*.f32 N/A

      \[\leadsto 1 + \color{blue}{v \cdot \log \left(u + \left(1 - u\right) \cdot e^{\frac{-2}{v}}\right)} \]

   2. *-commutative N/A

      \[\leadsto 1 + \color{blue}{\log \left(u + \left(1 - u\right) \cdot e^{\frac{-2}{v}}\right) \cdot v} \]

   3. lower-*.f32 99.6

      \[\leadsto 1 + \color{blue}{\log \left(u + \left(1 - u\right) \cdot e^{\frac{-2}{v}}\right) \cdot v} \]

   4. lift-+.f32 N/A

      \[\leadsto 1 + \log \color{blue}{\left(u + \left(1 - u\right) \cdot e^{\frac{-2}{v}}\right)} \cdot v \]

   5. +-commutative N/A

      \[\leadsto 1 + \log \color{blue}{\left(\left(1 - u\right) \cdot e^{\frac{-2}{v}} + u\right)} \cdot v \]

   6. lift-*.f32 N/A

      \[\leadsto 1 + \log \left(\color{blue}{\left(1 - u\right) \cdot e^{\frac{-2}{v}}} + u\right) \cdot v \]

   7. *-commutative N/A

      \[\leadsto 1 + \log \left(\color{blue}{e^{\frac{-2}{v}} \cdot \left(1 - u\right)} + u\right) \cdot v \]

   8. lower-fma.f32 99.6

      \[\leadsto 1 + \log \color{blue}{\left(\mathsf{fma}\left(e^{\frac{-2}{v}}, 1 - u, u\right)\right)} \cdot v \]

4. Applied rewrites 99.6%

   \[\leadsto 1 + \color{blue}{\log \left(\mathsf{fma}\left(e^{\frac{-2}{v}}, 1 - u, u\right)\right) \cdot v} \]
5. Step-by-step derivation

   1. lift-+.f32 N/A

      \[\leadsto \color{blue}{1 + \log \left(\mathsf{fma}\left(e^{\frac{-2}{v}}, 1 - u, u\right)\right) \cdot v} \]

   2. +-commutative N/A

      \[\leadsto \color{blue}{\log \left(\mathsf{fma}\left(e^{\frac{-2}{v}}, 1 - u, u\right)\right) \cdot v + 1} \]

   3. lift-*.f32 N/A

      \[\leadsto \color{blue}{\log \left(\mathsf{fma}\left(e^{\frac{-2}{v}}, 1 - u, u\right)\right) \cdot v} + 1 \]

   4. lower-fma.f32 99.6

      \[\leadsto \color{blue}{\mathsf{fma}\left(\log \left(\mathsf{fma}\left(e^{\frac{-2}{v}}, 1 - u, u\right)\right), v, 1\right)} \]

   5. lift-fma.f32 N/A

      \[\leadsto \mathsf{fma}\left(\log \color{blue}{\left(e^{\frac{-2}{v}} \cdot \left(1 - u\right) + u\right)}, v, 1\right) \]

   6. *-commutative N/A

      \[\leadsto \mathsf{fma}\left(\log \left(\color{blue}{\left(1 - u\right) \cdot e^{\frac{-2}{v}}} + u\right), v, 1\right) \]

   7. lower-fma.f32 99.6

      \[\leadsto \mathsf{fma}\left(\log \color{blue}{\left(\mathsf{fma}\left(1 - u, e^{\frac{-2}{v}}, u\right)\right)}, v, 1\right) \]

6. Applied rewrites 99.6%

   \[\leadsto \color{blue}{\mathsf{fma}\left(\log \left(\mathsf{fma}\left(1 - u, e^{\frac{-2}{v}}, u\right)\right), v, 1\right)} \]
7. Step-by-step derivation

   1. lift--.f32 N/A

      \[\leadsto \mathsf{fma}\left(\log \left(\mathsf{fma}\left(\color{blue}{1 - u}, e^{\frac{-2}{v}}, u\right)\right), v, 1\right) \]

   2. lift-fma.f32 N/A

      \[\leadsto \mathsf{fma}\left(\log \color{blue}{\left(\left(1 - u\right) \cdot e^{\frac{-2}{v}} + u\right)}, v, 1\right) \]

   3. lower-+.f32 N/A

      \[\leadsto \mathsf{fma}\left(\log \color{blue}{\left(\left(1 - u\right) \cdot e^{\frac{-2}{v}} + u\right)}, v, 1\right) \]

   4. lower-*.f32 N/A

      \[\leadsto \mathsf{fma}\left(\log \left(\color{blue}{\left(1 - u\right) \cdot e^{\frac{-2}{v}}} + u\right), v, 1\right) \]

   5. lift--.f32 99.6

      \[\leadsto \mathsf{fma}\left(\log \left(\color{blue}{\left(1 - u\right)} \cdot e^{\frac{-2}{v}} + u\right), v, 1\right) \]

8. Applied rewrites 99.6%

   \[\leadsto \mathsf{fma}\left(\log \color{blue}{\left(\left(1 - u\right) \cdot e^{\frac{-2}{v}} + u\right)}, v, 1\right) \]
9. Add Preprocessing

Alternative 3: 99.5% accurate, 1.0× speedup

Math

\[\begin{array}{l} \\ 1 + \log \left(\mathsf{fma}\left(e^{\frac{-2}{v}}, 1 - u, u\right)\right) \cdot v \end{array} \]

FPCore

(FPCore (u v)
 :precision binary32
 (+ 1.0 (* (log (fma (exp (/ -2.0 v)) (- 1.0 u) u)) v)))

C

float code(float u, float v) {
	return 1.0f + (logf(fmaf(expf((-2.0f / v)), (1.0f - u), u)) * v);
}

Julia

function code(u, v)
	return Float32(Float32(1.0) + Float32(log(fma(exp(Float32(Float32(-2.0) / v)), Float32(Float32(1.0) - u), u)) * v))
end

Derivation

1. Initial program 99.6%

   \[1 + v \cdot \log \left(u + \left(1 - u\right) \cdot e^{\frac{-2}{v}}\right) \]
2. Add Preprocessing
3. Step-by-step derivation

   1. lift-*.f32 N/A

      \[\leadsto 1 + \color{blue}{v \cdot \log \left(u + \left(1 - u\right) \cdot e^{\frac{-2}{v}}\right)} \]

   2. *-commutative N/A

      \[\leadsto 1 + \color{blue}{\log \left(u + \left(1 - u\right) \cdot e^{\frac{-2}{v}}\right) \cdot v} \]

   3. lower-*.f32 99.6

      \[\leadsto 1 + \color{blue}{\log \left(u + \left(1 - u\right) \cdot e^{\frac{-2}{v}}\right) \cdot v} \]

   4. lift-+.f32 N/A

      \[\leadsto 1 + \log \color{blue}{\left(u + \left(1 - u\right) \cdot e^{\frac{-2}{v}}\right)} \cdot v \]

   5. +-commutative N/A

      \[\leadsto 1 + \log \color{blue}{\left(\left(1 - u\right) \cdot e^{\frac{-2}{v}} + u\right)} \cdot v \]

   6. lift-*.f32 N/A

      \[\leadsto 1 + \log \left(\color{blue}{\left(1 - u\right) \cdot e^{\frac{-2}{v}}} + u\right) \cdot v \]

   7. *-commutative N/A

      \[\leadsto 1 + \log \left(\color{blue}{e^{\frac{-2}{v}} \cdot \left(1 - u\right)} + u\right) \cdot v \]

   8. lower-fma.f32 99.6

      \[\leadsto 1 + \log \color{blue}{\left(\mathsf{fma}\left(e^{\frac{-2}{v}}, 1 - u, u\right)\right)} \cdot v \]

4. Applied rewrites 99.6%

   \[\leadsto 1 + \color{blue}{\log \left(\mathsf{fma}\left(e^{\frac{-2}{v}}, 1 - u, u\right)\right) \cdot v} \]
5. Add Preprocessing

Alternative 4: 99.5% accurate, 1.0× speedup

Math

\[\begin{array}{l} \\ \mathsf{fma}\left(\log \left(\mathsf{fma}\left(e^{\frac{-2}{v}}, 1 - u, u\right)\right), v, 1\right) \end{array} \]

FPCore

(FPCore (u v)
 :precision binary32
 (fma (log (fma (exp (/ -2.0 v)) (- 1.0 u) u)) v 1.0))

C

float code(float u, float v) {
	return fmaf(logf(fmaf(expf((-2.0f / v)), (1.0f - u), u)), v, 1.0f);
}

Julia

function code(u, v)
	return fma(log(fma(exp(Float32(Float32(-2.0) / v)), Float32(Float32(1.0) - u), u)), v, Float32(1.0))
end

Derivation

1. Initial program 99.6%

   \[1 + v \cdot \log \left(u + \left(1 - u\right) \cdot e^{\frac{-2}{v}}\right) \]
2. Add Preprocessing
3. Step-by-step derivation

   1. lift-+.f32 N/A

      \[\leadsto \color{blue}{1 + v \cdot \log \left(u + \left(1 - u\right) \cdot e^{\frac{-2}{v}}\right)} \]

   2. +-commutative N/A

      \[\leadsto \color{blue}{v \cdot \log \left(u + \left(1 - u\right) \cdot e^{\frac{-2}{v}}\right) + 1} \]

   3. lift-*.f32 N/A

      \[\leadsto \color{blue}{v \cdot \log \left(u + \left(1 - u\right) \cdot e^{\frac{-2}{v}}\right)} + 1 \]

   4. *-commutative N/A

      \[\leadsto \color{blue}{\log \left(u + \left(1 - u\right) \cdot e^{\frac{-2}{v}}\right) \cdot v} + 1 \]

   5. lower-fma.f32 99.6

      \[\leadsto \color{blue}{\mathsf{fma}\left(\log \left(u + \left(1 - u\right) \cdot e^{\frac{-2}{v}}\right), v, 1\right)} \]

   6. lift-+.f32 N/A

      \[\leadsto \mathsf{fma}\left(\log \color{blue}{\left(u + \left(1 - u\right) \cdot e^{\frac{-2}{v}}\right)}, v, 1\right) \]

   7. +-commutative N/A

      \[\leadsto \mathsf{fma}\left(\log \color{blue}{\left(\left(1 - u\right) \cdot e^{\frac{-2}{v}} + u\right)}, v, 1\right) \]

   8. lift-*.f32 N/A

      \[\leadsto \mathsf{fma}\left(\log \left(\color{blue}{\left(1 - u\right) \cdot e^{\frac{-2}{v}}} + u\right), v, 1\right) \]

   9. *-commutative N/A

      \[\leadsto \mathsf{fma}\left(\log \left(\color{blue}{e^{\frac{-2}{v}} \cdot \left(1 - u\right)} + u\right), v, 1\right) \]

   10. lower-fma.f32 99.6

       \[\leadsto \mathsf{fma}\left(\log \color{blue}{\left(\mathsf{fma}\left(e^{\frac{-2}{v}}, 1 - u, u\right)\right)}, v, 1\right) \]

4. Applied rewrites 99.6%

   \[\leadsto \color{blue}{\mathsf{fma}\left(\log \left(\mathsf{fma}\left(e^{\frac{-2}{v}}, 1 - u, u\right)\right), v, 1\right)} \]
5. Add Preprocessing

Alternative 5: 95.8% accurate, 1.0× speedup

Math

\[\begin{array}{l} \\ 1 + \log \left(\mathsf{fma}\left(1, e^{\frac{-2}{v}}, u\right)\right) \cdot v \end{array} \]

FPCore

(FPCore (u v)
 :precision binary32
 (+ 1.0 (* (log (fma 1.0 (exp (/ -2.0 v)) u)) v)))

C

float code(float u, float v) {
	return 1.0f + (logf(fmaf(1.0f, expf((-2.0f / v)), u)) * v);
}

Julia

function code(u, v)
	return Float32(Float32(1.0) + Float32(log(fma(Float32(1.0), exp(Float32(Float32(-2.0) / v)), u)) * v))
end

Derivation

1. Initial program 99.6%

   \[1 + v \cdot \log \left(u + \left(1 - u\right) \cdot e^{\frac{-2}{v}}\right) \]
2. Add Preprocessing

3. Taylor expanded in u around 0

   \[\leadsto 1 + v \cdot \log \left(u + \color{blue}{1} \cdot e^{\frac{-2}{v}}\right) \]
4. Step-by-step derivation

5. Applied rewrites 97.4%

   \[\leadsto 1 + v \cdot \log \left(u + \color{blue}{1} \cdot e^{\frac{-2}{v}}\right) \]
6. Step-by-step derivation

   1. lift-*.f32 N/A

      \[\leadsto 1 + \color{blue}{v \cdot \log \left(u + 1 \cdot e^{\frac{-2}{v}}\right)} \]

   2. *-commutative N/A

      \[\leadsto 1 + \color{blue}{\log \left(u + 1 \cdot e^{\frac{-2}{v}}\right) \cdot v} \]

   3. lower-*.f32 97.4

      \[\leadsto 1 + \color{blue}{\log \left(u + 1 \cdot e^{\frac{-2}{v}}\right) \cdot v} \]

   4. lift-+.f32 N/A

      \[\leadsto 1 + \log \color{blue}{\left(u + 1 \cdot e^{\frac{-2}{v}}\right)} \cdot v \]

   5. +-commutative N/A

      \[\leadsto 1 + \log \color{blue}{\left(1 \cdot e^{\frac{-2}{v}} + u\right)} \cdot v \]

   6. lift-*.f32 N/A

      \[\leadsto 1 + \log \left(\color{blue}{1 \cdot e^{\frac{-2}{v}}} + u\right) \cdot v \]

   7. lower-fma.f32 97.4

      \[\leadsto 1 + \log \color{blue}{\left(\mathsf{fma}\left(1, e^{\frac{-2}{v}}, u\right)\right)} \cdot v \]

7. Applied rewrites 97.4%

   \[\leadsto 1 + \color{blue}{\log \left(\mathsf{fma}\left(1, e^{\frac{-2}{v}}, u\right)\right) \cdot v} \]
8. Add Preprocessing

Alternative 6: 95.8% accurate, 1.0× speedup

Math

\[\begin{array}{l} \\ \mathsf{fma}\left(\log \left(\mathsf{fma}\left(1, e^{\frac{-2}{v}}, u\right)\right), v, 1\right) \end{array} \]

FPCore

(FPCore (u v)
 :precision binary32
 (fma (log (fma 1.0 (exp (/ -2.0 v)) u)) v 1.0))

C

float code(float u, float v) {
	return fmaf(logf(fmaf(1.0f, expf((-2.0f / v)), u)), v, 1.0f);
}

Julia

function code(u, v)
	return fma(log(fma(Float32(1.0), exp(Float32(Float32(-2.0) / v)), u)), v, Float32(1.0))
end

Derivation

1. Initial program 99.6%

   \[1 + v \cdot \log \left(u + \left(1 - u\right) \cdot e^{\frac{-2}{v}}\right) \]
2. Add Preprocessing

3. Taylor expanded in u around 0

   \[\leadsto 1 + v \cdot \log \left(u + \color{blue}{1} \cdot e^{\frac{-2}{v}}\right) \]
4. Step-by-step derivation

5. Applied rewrites 97.4%

   \[\leadsto 1 + v \cdot \log \left(u + \color{blue}{1} \cdot e^{\frac{-2}{v}}\right) \]
6. Step-by-step derivation

   1. lift-+.f32 N/A

      \[\leadsto \color{blue}{1 + v \cdot \log \left(u + 1 \cdot e^{\frac{-2}{v}}\right)} \]

   2. +-commutative N/A

      \[\leadsto \color{blue}{v \cdot \log \left(u + 1 \cdot e^{\frac{-2}{v}}\right) + 1} \]

   3. lift-*.f32 N/A

      \[\leadsto \color{blue}{v \cdot \log \left(u + 1 \cdot e^{\frac{-2}{v}}\right)} + 1 \]

   4. *-commutative N/A

      \[\leadsto \color{blue}{\log \left(u + 1 \cdot e^{\frac{-2}{v}}\right) \cdot v} + 1 \]

   5. lower-fma.f32 97.3

      \[\leadsto \color{blue}{\mathsf{fma}\left(\log \left(u + 1 \cdot e^{\frac{-2}{v}}\right), v, 1\right)} \]

   6. lift-+.f32 N/A

      \[\leadsto \mathsf{fma}\left(\log \color{blue}{\left(u + 1 \cdot e^{\frac{-2}{v}}\right)}, v, 1\right) \]

   7. +-commutative N/A

      \[\leadsto \mathsf{fma}\left(\log \color{blue}{\left(1 \cdot e^{\frac{-2}{v}} + u\right)}, v, 1\right) \]

   8. lift-*.f32 N/A

      \[\leadsto \mathsf{fma}\left(\log \left(\color{blue}{1 \cdot e^{\frac{-2}{v}}} + u\right), v, 1\right) \]

   9. lower-fma.f32 97.3

      \[\leadsto \mathsf{fma}\left(\log \color{blue}{\left(\mathsf{fma}\left(1, e^{\frac{-2}{v}}, u\right)\right)}, v, 1\right) \]

7. Applied rewrites 97.3%

   \[\leadsto \color{blue}{\mathsf{fma}\left(\log \left(\mathsf{fma}\left(1, e^{\frac{-2}{v}}, u\right)\right), v, 1\right)} \]
8. Add Preprocessing

Alternative 7: 87.2% accurate, 2.0× speedup

Math

\[\begin{array}{l} \\ 1 + v \cdot \log \left(u + \left(1 - u\right)\right) \end{array} \]

FPCore

(FPCore (u v) :precision binary32 (+ 1.0 (* v (log (+ u (- 1.0 u))))))

C

float code(float u, float v) {
	return 1.0f + (v * logf((u + (1.0f - u))));
}

Fortran

module fmin_fmax_functions
    implicit none
    private
    public fmax
    public fmin

    interface fmax
        module procedure fmax88
        module procedure fmax44
        module procedure fmax84
        module procedure fmax48
    end interface
    interface fmin
        module procedure fmin88
        module procedure fmin44
        module procedure fmin84
        module procedure fmin48
    end interface
contains
    real(8) function fmax88(x, y) result (res)
        real(8), intent (in) :: x
        real(8), intent (in) :: y
        res = merge(y, merge(x, max(x, y), y /= y), x /= x)
    end function
    real(4) function fmax44(x, y) result (res)
        real(4), intent (in) :: x
        real(4), intent (in) :: y
        res = merge(y, merge(x, max(x, y), y /= y), x /= x)
    end function
    real(8) function fmax84(x, y) result(res)
        real(8), intent (in) :: x
        real(4), intent (in) :: y
        res = merge(dble(y), merge(x, max(x, dble(y)), y /= y), x /= x)
    end function
    real(8) function fmax48(x, y) result(res)
        real(4), intent (in) :: x
        real(8), intent (in) :: y
        res = merge(y, merge(dble(x), max(dble(x), y), y /= y), x /= x)
    end function
    real(8) function fmin88(x, y) result (res)
        real(8), intent (in) :: x
        real(8), intent (in) :: y
        res = merge(y, merge(x, min(x, y), y /= y), x /= x)
    end function
    real(4) function fmin44(x, y) result (res)
        real(4), intent (in) :: x
        real(4), intent (in) :: y
        res = merge(y, merge(x, min(x, y), y /= y), x /= x)
    end function
    real(8) function fmin84(x, y) result(res)
        real(8), intent (in) :: x
        real(4), intent (in) :: y
        res = merge(dble(y), merge(x, min(x, dble(y)), y /= y), x /= x)
    end function
    real(8) function fmin48(x, y) result(res)
        real(4), intent (in) :: x
        real(8), intent (in) :: y
        res = merge(y, merge(dble(x), min(dble(x), y), y /= y), x /= x)
    end function
end module

real(4) function code(u, v)
use fmin_fmax_functions
    real(4), intent (in) :: u
    real(4), intent (in) :: v
    code = 1.0e0 + (v * log((u + (1.0e0 - u))))
end function

Julia

function code(u, v)
	return Float32(Float32(1.0) + Float32(v * log(Float32(u + Float32(Float32(1.0) - u)))))
end

MATLAB

function tmp = code(u, v)
	tmp = single(1.0) + (v * log((u + (single(1.0) - u))));
end

Derivation

1. Initial program 99.6%

   \[1 + v \cdot \log \left(u + \left(1 - u\right) \cdot e^{\frac{-2}{v}}\right) \]
2. Add Preprocessing

3. Taylor expanded in v around inf

   \[\leadsto 1 + v \cdot \log \left(u + \color{blue}{\left(1 - u\right)}\right) \]
4. Step-by-step derivation

   1. lower--.f32 87.9

      \[\leadsto 1 + v \cdot \log \left(u + \color{blue}{\left(1 - u\right)}\right) \]

5. Applied rewrites 87.9%

   \[\leadsto 1 + v \cdot \log \left(u + \color{blue}{\left(1 - u\right)}\right) \]
6. Add Preprocessing

Alternative 8: 20.4% accurate, 4.3× speedup

Math

\[\begin{array}{l} \\ \left(\frac{\frac{0.5}{u} - 1}{u} + 2\right) \cdot u - \frac{1}{\mathsf{fma}\left(2, u, 1\right)} \end{array} \]

FPCore

(FPCore (u v)
 :precision binary32
 (- (* (+ (/ (- (/ 0.5 u) 1.0) u) 2.0) u) (/ 1.0 (fma 2.0 u 1.0))))

C

float code(float u, float v) {
	return (((((0.5f / u) - 1.0f) / u) + 2.0f) * u) - (1.0f / fmaf(2.0f, u, 1.0f));
}

Julia

function code(u, v)
	return Float32(Float32(Float32(Float32(Float32(Float32(Float32(0.5) / u) - Float32(1.0)) / u) + Float32(2.0)) * u) - Float32(Float32(1.0) / fma(Float32(2.0), u, Float32(1.0))))
end

Derivation

1. Initial program 99.6%

   \[1 + v \cdot \log \left(u + \left(1 - u\right) \cdot e^{\frac{-2}{v}}\right) \]
2. Add Preprocessing

3. Taylor expanded in u around 0

   \[\leadsto \color{blue}{u \cdot \left(v \cdot \left(\frac{1}{e^{\frac{-2}{v}}} - 1\right)\right) - 1} \]
4. Step-by-step derivation

   1. metadata-eval N/A

      \[\leadsto u \cdot \left(v \cdot \left(\frac{1}{e^{\frac{-2}{v}}} - 1\right)\right) - \color{blue}{1 \cdot 1} \]

   2. fp-cancel-sub-sign-inv N/A

      \[\leadsto \color{blue}{u \cdot \left(v \cdot \left(\frac{1}{e^{\frac{-2}{v}}} - 1\right)\right) + \left(\mathsf{neg}\left(1\right)\right) \cdot 1} \]

   3. associate-*r* N/A

      \[\leadsto \color{blue}{\left(u \cdot v\right) \cdot \left(\frac{1}{e^{\frac{-2}{v}}} - 1\right)} + \left(\mathsf{neg}\left(1\right)\right) \cdot 1 \]

   4. metadata-eval N/A

      \[\leadsto \left(u \cdot v\right) \cdot \left(\frac{1}{e^{\frac{-2}{v}}} - 1\right) + \color{blue}{-1} \cdot 1 \]

   5. metadata-eval N/A

      \[\leadsto \left(u \cdot v\right) \cdot \left(\frac{1}{e^{\frac{-2}{v}}} - 1\right) + \color{blue}{-1} \]

   6. lower-fma.f32 N/A

      \[\leadsto \color{blue}{\mathsf{fma}\left(u \cdot v, \frac{1}{e^{\frac{-2}{v}}} - 1, -1\right)} \]

   7. lower-*.f32 N/A

      \[\leadsto \mathsf{fma}\left(\color{blue}{u \cdot v}, \frac{1}{e^{\frac{-2}{v}}} - 1, -1\right) \]

   8. rec-exp N/A

      \[\leadsto \mathsf{fma}\left(u \cdot v, \color{blue}{e^{\mathsf{neg}\left(\frac{-2}{v}\right)}} - 1, -1\right) \]

   9. distribute-frac-neg N/A

      \[\leadsto \mathsf{fma}\left(u \cdot v, e^{\color{blue}{\frac{\mathsf{neg}\left(-2\right)}{v}}} - 1, -1\right) \]

   10. metadata-eval N/A

       \[\leadsto \mathsf{fma}\left(u \cdot v, e^{\frac{\color{blue}{2}}{v}} - 1, -1\right) \]

   11. metadata-eval N/A

       \[\leadsto \mathsf{fma}\left(u \cdot v, e^{\frac{\color{blue}{2 \cdot 1}}{v}} - 1, -1\right) \]

   12. associate-*r/ N/A

       \[\leadsto \mathsf{fma}\left(u \cdot v, e^{\color{blue}{2 \cdot \frac{1}{v}}} - 1, -1\right) \]

   13. lower-expm1.f32 N/A

       \[\leadsto \mathsf{fma}\left(u \cdot v, \color{blue}{\mathsf{expm1}\left(2 \cdot \frac{1}{v}\right)}, -1\right) \]

   14. associate-*r/ N/A

       \[\leadsto \mathsf{fma}\left(u \cdot v, \mathsf{expm1}\left(\color{blue}{\frac{2 \cdot 1}{v}}\right), -1\right) \]

   15. metadata-eval N/A

       \[\leadsto \mathsf{fma}\left(u \cdot v, \mathsf{expm1}\left(\frac{\color{blue}{2}}{v}\right), -1\right) \]

   16. lower-/.f32 9.4

       \[\leadsto \mathsf{fma}\left(u \cdot v, \mathsf{expm1}\left(\color{blue}{\frac{2}{v}}\right), -1\right) \]

5. Applied rewrites 9.4%

   \[\leadsto \color{blue}{\mathsf{fma}\left(u \cdot v, \mathsf{expm1}\left(\frac{2}{v}\right), -1\right)} \]

6. Taylor expanded in v around inf

   \[\leadsto 2 \cdot u - \color{blue}{1} \]
7. Step-by-step derivation

8. Applied rewrites 7.4%

   \[\leadsto 2 \cdot u - \color{blue}{1} \]
9. Step-by-step derivation

10. Applied rewrites 7.4%

    \[\leadsto \frac{4 \cdot \left(u \cdot u\right)}{\mathsf{fma}\left(2, u, 1\right)} - \frac{1}{\color{blue}{\mathsf{fma}\left(2, u, 1\right)}} \]

11. Taylor expanded in u around inf

    \[\leadsto u \cdot \left(\left(2 + \frac{\frac{1}{2}}{{u}^{2}}\right) - \frac{1}{u}\right) - \frac{1}{\mathsf{fma}\left(\color{blue}{2}, u, 1\right)} \]
12. Step-by-step derivation

13. Applied rewrites 21.0%

    \[\leadsto \left(\frac{\frac{0.5}{u} - 1}{u} + 2\right) \cdot u - \frac{1}{\mathsf{fma}\left(\color{blue}{2}, u, 1\right)} \]
14. Add Preprocessing

Alternative 9: 14.2% accurate, 4.9× speedup

Math

\[\begin{array}{l} \\ 1 + \mathsf{fma}\left(1 - u, -2, \frac{\mathsf{fma}\left(\mathsf{fma}\left(u - 2, u, 1\right), -2, \left(1 - u\right) \cdot 2\right)}{v}\right) \end{array} \]

FPCore

(FPCore (u v)
 :precision binary32
 (+
  1.0
  (fma
   (- 1.0 u)
   -2.0
   (/ (fma (fma (- u 2.0) u 1.0) -2.0 (* (- 1.0 u) 2.0)) v))))

C

float code(float u, float v) {
	return 1.0f + fmaf((1.0f - u), -2.0f, (fmaf(fmaf((u - 2.0f), u, 1.0f), -2.0f, ((1.0f - u) * 2.0f)) / v));
}

Julia

function code(u, v)
	return Float32(Float32(1.0) + fma(Float32(Float32(1.0) - u), Float32(-2.0), Float32(fma(fma(Float32(u - Float32(2.0)), u, Float32(1.0)), Float32(-2.0), Float32(Float32(Float32(1.0) - u) * Float32(2.0))) / v)))
end

Derivation

1. Initial program 99.6%

   \[1 + v \cdot \log \left(u + \left(1 - u\right) \cdot e^{\frac{-2}{v}}\right) \]
2. Add Preprocessing

3. Taylor expanded in v around inf

   \[\leadsto 1 + \color{blue}{\left(-2 \cdot \left(1 - u\right) + \frac{1}{2} \cdot \frac{-4 \cdot {\left(1 - u\right)}^{2} + 4 \cdot \left(1 - u\right)}{v}\right)} \]
4. Step-by-step derivation

   1. *-commutative N/A

      \[\leadsto 1 + \left(\color{blue}{\left(1 - u\right) \cdot -2} + \frac{1}{2} \cdot \frac{-4 \cdot {\left(1 - u\right)}^{2} + 4 \cdot \left(1 - u\right)}{v}\right) \]

   2. lower-fma.f32 N/A

      \[\leadsto 1 + \color{blue}{\mathsf{fma}\left(1 - u, -2, \frac{1}{2} \cdot \frac{-4 \cdot {\left(1 - u\right)}^{2} + 4 \cdot \left(1 - u\right)}{v}\right)} \]

   3. lower--.f32 N/A

      \[\leadsto 1 + \mathsf{fma}\left(\color{blue}{1 - u}, -2, \frac{1}{2} \cdot \frac{-4 \cdot {\left(1 - u\right)}^{2} + 4 \cdot \left(1 - u\right)}{v}\right) \]

   4. associate-*r/ N/A

      \[\leadsto 1 + \mathsf{fma}\left(1 - u, -2, \color{blue}{\frac{\frac{1}{2} \cdot \left(-4 \cdot {\left(1 - u\right)}^{2} + 4 \cdot \left(1 - u\right)\right)}{v}}\right) \]

   5. lower-/.f32 N/A

      \[\leadsto 1 + \mathsf{fma}\left(1 - u, -2, \color{blue}{\frac{\frac{1}{2} \cdot \left(-4 \cdot {\left(1 - u\right)}^{2} + 4 \cdot \left(1 - u\right)\right)}{v}}\right) \]

   6. distribute-lft-in N/A

      \[\leadsto 1 + \mathsf{fma}\left(1 - u, -2, \frac{\color{blue}{\frac{1}{2} \cdot \left(-4 \cdot {\left(1 - u\right)}^{2}\right) + \frac{1}{2} \cdot \left(4 \cdot \left(1 - u\right)\right)}}{v}\right) \]

   7. *-commutative N/A

      \[\leadsto 1 + \mathsf{fma}\left(1 - u, -2, \frac{\color{blue}{\left(-4 \cdot {\left(1 - u\right)}^{2}\right) \cdot \frac{1}{2}} + \frac{1}{2} \cdot \left(4 \cdot \left(1 - u\right)\right)}{v}\right) \]

   8. *-commutative N/A

      \[\leadsto 1 + \mathsf{fma}\left(1 - u, -2, \frac{\color{blue}{\left({\left(1 - u\right)}^{2} \cdot -4\right)} \cdot \frac{1}{2} + \frac{1}{2} \cdot \left(4 \cdot \left(1 - u\right)\right)}{v}\right) \]

   9. associate-*l* N/A

      \[\leadsto 1 + \mathsf{fma}\left(1 - u, -2, \frac{\color{blue}{{\left(1 - u\right)}^{2} \cdot \left(-4 \cdot \frac{1}{2}\right)} + \frac{1}{2} \cdot \left(4 \cdot \left(1 - u\right)\right)}{v}\right) \]

   10. metadata-eval N/A

       \[\leadsto 1 + \mathsf{fma}\left(1 - u, -2, \frac{{\left(1 - u\right)}^{2} \cdot \color{blue}{-2} + \frac{1}{2} \cdot \left(4 \cdot \left(1 - u\right)\right)}{v}\right) \]

   11. associate-*r* N/A

       \[\leadsto 1 + \mathsf{fma}\left(1 - u, -2, \frac{{\left(1 - u\right)}^{2} \cdot -2 + \color{blue}{\left(\frac{1}{2} \cdot 4\right) \cdot \left(1 - u\right)}}{v}\right) \]

   12. metadata-eval N/A

       \[\leadsto 1 + \mathsf{fma}\left(1 - u, -2, \frac{{\left(1 - u\right)}^{2} \cdot -2 + \color{blue}{2} \cdot \left(1 - u\right)}{v}\right) \]

   13. lower-fma.f32 N/A

       \[\leadsto 1 + \mathsf{fma}\left(1 - u, -2, \frac{\color{blue}{\mathsf{fma}\left({\left(1 - u\right)}^{2}, -2, 2 \cdot \left(1 - u\right)\right)}}{v}\right) \]

   14. lower-pow.f32 N/A

       \[\leadsto 1 + \mathsf{fma}\left(1 - u, -2, \frac{\mathsf{fma}\left(\color{blue}{{\left(1 - u\right)}^{2}}, -2, 2 \cdot \left(1 - u\right)\right)}{v}\right) \]

   15. lower--.f32 N/A

       \[\leadsto 1 + \mathsf{fma}\left(1 - u, -2, \frac{\mathsf{fma}\left({\color{blue}{\left(1 - u\right)}}^{2}, -2, 2 \cdot \left(1 - u\right)\right)}{v}\right) \]

   16. *-commutative N/A

       \[\leadsto 1 + \mathsf{fma}\left(1 - u, -2, \frac{\mathsf{fma}\left({\left(1 - u\right)}^{2}, -2, \color{blue}{\left(1 - u\right) \cdot 2}\right)}{v}\right) \]

   17. lower-*.f32 N/A

       \[\leadsto 1 + \mathsf{fma}\left(1 - u, -2, \frac{\mathsf{fma}\left({\left(1 - u\right)}^{2}, -2, \color{blue}{\left(1 - u\right) \cdot 2}\right)}{v}\right) \]

   18. lower--.f32 13.4

       \[\leadsto 1 + \mathsf{fma}\left(1 - u, -2, \frac{\mathsf{fma}\left({\left(1 - u\right)}^{2}, -2, \color{blue}{\left(1 - u\right)} \cdot 2\right)}{v}\right) \]

5. Applied rewrites 13.4%

   \[\leadsto 1 + \color{blue}{\mathsf{fma}\left(1 - u, -2, \frac{\mathsf{fma}\left({\left(1 - u\right)}^{2}, -2, \left(1 - u\right) \cdot 2\right)}{v}\right)} \]

6. Taylor expanded in u around 0

   \[\leadsto 1 + \mathsf{fma}\left(1 - u, -2, \frac{\mathsf{fma}\left(1 + u \cdot \left(u - 2\right), -2, \left(1 - u\right) \cdot 2\right)}{v}\right) \]
7. Step-by-step derivation

8. Applied rewrites 13.4%

   \[\leadsto 1 + \mathsf{fma}\left(1 - u, -2, \frac{\mathsf{fma}\left(\mathsf{fma}\left(u - 2, u, 1\right), -2, \left(1 - u\right) \cdot 2\right)}{v}\right) \]
9. Add Preprocessing

Alternative 10: 14.2% accurate, 6.6× speedup

Math

\[\begin{array}{l} \\ 1 + \mathsf{fma}\left(1 - u, -2, \frac{\mathsf{fma}\left(-2, u, 2\right) \cdot u}{v}\right) \end{array} \]

FPCore

(FPCore (u v)
 :precision binary32
 (+ 1.0 (fma (- 1.0 u) -2.0 (/ (* (fma -2.0 u 2.0) u) v))))

C

float code(float u, float v) {
	return 1.0f + fmaf((1.0f - u), -2.0f, ((fmaf(-2.0f, u, 2.0f) * u) / v));
}

Julia

function code(u, v)
	return Float32(Float32(1.0) + fma(Float32(Float32(1.0) - u), Float32(-2.0), Float32(Float32(fma(Float32(-2.0), u, Float32(2.0)) * u) / v)))
end

Derivation

1. Initial program 99.6%

   \[1 + v \cdot \log \left(u + \left(1 - u\right) \cdot e^{\frac{-2}{v}}\right) \]
2. Add Preprocessing

3. Taylor expanded in v around inf

   \[\leadsto 1 + \color{blue}{\left(-2 \cdot \left(1 - u\right) + \frac{1}{2} \cdot \frac{-4 \cdot {\left(1 - u\right)}^{2} + 4 \cdot \left(1 - u\right)}{v}\right)} \]
4. Step-by-step derivation

   1. *-commutative N/A

      \[\leadsto 1 + \left(\color{blue}{\left(1 - u\right) \cdot -2} + \frac{1}{2} \cdot \frac{-4 \cdot {\left(1 - u\right)}^{2} + 4 \cdot \left(1 - u\right)}{v}\right) \]

   2. lower-fma.f32 N/A

      \[\leadsto 1 + \color{blue}{\mathsf{fma}\left(1 - u, -2, \frac{1}{2} \cdot \frac{-4 \cdot {\left(1 - u\right)}^{2} + 4 \cdot \left(1 - u\right)}{v}\right)} \]

   3. lower--.f32 N/A

      \[\leadsto 1 + \mathsf{fma}\left(\color{blue}{1 - u}, -2, \frac{1}{2} \cdot \frac{-4 \cdot {\left(1 - u\right)}^{2} + 4 \cdot \left(1 - u\right)}{v}\right) \]

   4. associate-*r/ N/A

      \[\leadsto 1 + \mathsf{fma}\left(1 - u, -2, \color{blue}{\frac{\frac{1}{2} \cdot \left(-4 \cdot {\left(1 - u\right)}^{2} + 4 \cdot \left(1 - u\right)\right)}{v}}\right) \]

   5. lower-/.f32 N/A

      \[\leadsto 1 + \mathsf{fma}\left(1 - u, -2, \color{blue}{\frac{\frac{1}{2} \cdot \left(-4 \cdot {\left(1 - u\right)}^{2} + 4 \cdot \left(1 - u\right)\right)}{v}}\right) \]

   6. distribute-lft-in N/A

      \[\leadsto 1 + \mathsf{fma}\left(1 - u, -2, \frac{\color{blue}{\frac{1}{2} \cdot \left(-4 \cdot {\left(1 - u\right)}^{2}\right) + \frac{1}{2} \cdot \left(4 \cdot \left(1 - u\right)\right)}}{v}\right) \]

   7. *-commutative N/A

      \[\leadsto 1 + \mathsf{fma}\left(1 - u, -2, \frac{\color{blue}{\left(-4 \cdot {\left(1 - u\right)}^{2}\right) \cdot \frac{1}{2}} + \frac{1}{2} \cdot \left(4 \cdot \left(1 - u\right)\right)}{v}\right) \]

   8. *-commutative N/A

      \[\leadsto 1 + \mathsf{fma}\left(1 - u, -2, \frac{\color{blue}{\left({\left(1 - u\right)}^{2} \cdot -4\right)} \cdot \frac{1}{2} + \frac{1}{2} \cdot \left(4 \cdot \left(1 - u\right)\right)}{v}\right) \]

   9. associate-*l* N/A

      \[\leadsto 1 + \mathsf{fma}\left(1 - u, -2, \frac{\color{blue}{{\left(1 - u\right)}^{2} \cdot \left(-4 \cdot \frac{1}{2}\right)} + \frac{1}{2} \cdot \left(4 \cdot \left(1 - u\right)\right)}{v}\right) \]

   10. metadata-eval N/A

       \[\leadsto 1 + \mathsf{fma}\left(1 - u, -2, \frac{{\left(1 - u\right)}^{2} \cdot \color{blue}{-2} + \frac{1}{2} \cdot \left(4 \cdot \left(1 - u\right)\right)}{v}\right) \]

   11. associate-*r* N/A

       \[\leadsto 1 + \mathsf{fma}\left(1 - u, -2, \frac{{\left(1 - u\right)}^{2} \cdot -2 + \color{blue}{\left(\frac{1}{2} \cdot 4\right) \cdot \left(1 - u\right)}}{v}\right) \]

   12. metadata-eval N/A

       \[\leadsto 1 + \mathsf{fma}\left(1 - u, -2, \frac{{\left(1 - u\right)}^{2} \cdot -2 + \color{blue}{2} \cdot \left(1 - u\right)}{v}\right) \]

   13. lower-fma.f32 N/A

       \[\leadsto 1 + \mathsf{fma}\left(1 - u, -2, \frac{\color{blue}{\mathsf{fma}\left({\left(1 - u\right)}^{2}, -2, 2 \cdot \left(1 - u\right)\right)}}{v}\right) \]

   14. lower-pow.f32 N/A

       \[\leadsto 1 + \mathsf{fma}\left(1 - u, -2, \frac{\mathsf{fma}\left(\color{blue}{{\left(1 - u\right)}^{2}}, -2, 2 \cdot \left(1 - u\right)\right)}{v}\right) \]

   15. lower--.f32 N/A

       \[\leadsto 1 + \mathsf{fma}\left(1 - u, -2, \frac{\mathsf{fma}\left({\color{blue}{\left(1 - u\right)}}^{2}, -2, 2 \cdot \left(1 - u\right)\right)}{v}\right) \]

   16. *-commutative N/A

       \[\leadsto 1 + \mathsf{fma}\left(1 - u, -2, \frac{\mathsf{fma}\left({\left(1 - u\right)}^{2}, -2, \color{blue}{\left(1 - u\right) \cdot 2}\right)}{v}\right) \]

   17. lower-*.f32 N/A

       \[\leadsto 1 + \mathsf{fma}\left(1 - u, -2, \frac{\mathsf{fma}\left({\left(1 - u\right)}^{2}, -2, \color{blue}{\left(1 - u\right) \cdot 2}\right)}{v}\right) \]

   18. lower--.f32 13.4

       \[\leadsto 1 + \mathsf{fma}\left(1 - u, -2, \frac{\mathsf{fma}\left({\left(1 - u\right)}^{2}, -2, \color{blue}{\left(1 - u\right)} \cdot 2\right)}{v}\right) \]

5. Applied rewrites 13.4%

   \[\leadsto 1 + \color{blue}{\mathsf{fma}\left(1 - u, -2, \frac{\mathsf{fma}\left({\left(1 - u\right)}^{2}, -2, \left(1 - u\right) \cdot 2\right)}{v}\right)} \]

6. Taylor expanded in u around 0

   \[\leadsto 1 + \mathsf{fma}\left(1 - u, -2, \frac{u \cdot \left(2 + -2 \cdot u\right)}{v}\right) \]
7. Step-by-step derivation

8. Applied rewrites 13.4%

   \[\leadsto 1 + \mathsf{fma}\left(1 - u, -2, \frac{\mathsf{fma}\left(-2, u, 2\right) \cdot u}{v}\right) \]
9. Add Preprocessing

Alternative 11: 14.2% accurate, 7.0× speedup

Math

\[\begin{array}{l} \\ 1 + \mathsf{fma}\left(\frac{\mathsf{fma}\left(-2, u, 1\right) + 1}{v} - -2, u, -2\right) \end{array} \]

FPCore

(FPCore (u v)
 :precision binary32
 (+ 1.0 (fma (- (/ (+ (fma -2.0 u 1.0) 1.0) v) -2.0) u -2.0)))

C

float code(float u, float v) {
	return 1.0f + fmaf((((fmaf(-2.0f, u, 1.0f) + 1.0f) / v) - -2.0f), u, -2.0f);
}

Julia

function code(u, v)
	return Float32(Float32(1.0) + fma(Float32(Float32(Float32(fma(Float32(-2.0), u, Float32(1.0)) + Float32(1.0)) / v) - Float32(-2.0)), u, Float32(-2.0)))
end

Derivation

1. Initial program 99.6%

   \[1 + v \cdot \log \left(u + \left(1 - u\right) \cdot e^{\frac{-2}{v}}\right) \]
2. Add Preprocessing

3. Taylor expanded in v around inf

   \[\leadsto 1 + \color{blue}{\left(-2 \cdot \left(1 - u\right) + \frac{1}{2} \cdot \frac{-4 \cdot {\left(1 - u\right)}^{2} + 4 \cdot \left(1 - u\right)}{v}\right)} \]
4. Step-by-step derivation

   1. *-commutative N/A

      \[\leadsto 1 + \left(\color{blue}{\left(1 - u\right) \cdot -2} + \frac{1}{2} \cdot \frac{-4 \cdot {\left(1 - u\right)}^{2} + 4 \cdot \left(1 - u\right)}{v}\right) \]

   2. lower-fma.f32 N/A

      \[\leadsto 1 + \color{blue}{\mathsf{fma}\left(1 - u, -2, \frac{1}{2} \cdot \frac{-4 \cdot {\left(1 - u\right)}^{2} + 4 \cdot \left(1 - u\right)}{v}\right)} \]

   3. lower--.f32 N/A

      \[\leadsto 1 + \mathsf{fma}\left(\color{blue}{1 - u}, -2, \frac{1}{2} \cdot \frac{-4 \cdot {\left(1 - u\right)}^{2} + 4 \cdot \left(1 - u\right)}{v}\right) \]

   4. associate-*r/ N/A

      \[\leadsto 1 + \mathsf{fma}\left(1 - u, -2, \color{blue}{\frac{\frac{1}{2} \cdot \left(-4 \cdot {\left(1 - u\right)}^{2} + 4 \cdot \left(1 - u\right)\right)}{v}}\right) \]

   5. lower-/.f32 N/A

      \[\leadsto 1 + \mathsf{fma}\left(1 - u, -2, \color{blue}{\frac{\frac{1}{2} \cdot \left(-4 \cdot {\left(1 - u\right)}^{2} + 4 \cdot \left(1 - u\right)\right)}{v}}\right) \]

   6. distribute-lft-in N/A

      \[\leadsto 1 + \mathsf{fma}\left(1 - u, -2, \frac{\color{blue}{\frac{1}{2} \cdot \left(-4 \cdot {\left(1 - u\right)}^{2}\right) + \frac{1}{2} \cdot \left(4 \cdot \left(1 - u\right)\right)}}{v}\right) \]

   7. *-commutative N/A

      \[\leadsto 1 + \mathsf{fma}\left(1 - u, -2, \frac{\color{blue}{\left(-4 \cdot {\left(1 - u\right)}^{2}\right) \cdot \frac{1}{2}} + \frac{1}{2} \cdot \left(4 \cdot \left(1 - u\right)\right)}{v}\right) \]

   8. *-commutative N/A

      \[\leadsto 1 + \mathsf{fma}\left(1 - u, -2, \frac{\color{blue}{\left({\left(1 - u\right)}^{2} \cdot -4\right)} \cdot \frac{1}{2} + \frac{1}{2} \cdot \left(4 \cdot \left(1 - u\right)\right)}{v}\right) \]

   9. associate-*l* N/A

      \[\leadsto 1 + \mathsf{fma}\left(1 - u, -2, \frac{\color{blue}{{\left(1 - u\right)}^{2} \cdot \left(-4 \cdot \frac{1}{2}\right)} + \frac{1}{2} \cdot \left(4 \cdot \left(1 - u\right)\right)}{v}\right) \]

   10. metadata-eval N/A

       \[\leadsto 1 + \mathsf{fma}\left(1 - u, -2, \frac{{\left(1 - u\right)}^{2} \cdot \color{blue}{-2} + \frac{1}{2} \cdot \left(4 \cdot \left(1 - u\right)\right)}{v}\right) \]

   11. associate-*r* N/A

       \[\leadsto 1 + \mathsf{fma}\left(1 - u, -2, \frac{{\left(1 - u\right)}^{2} \cdot -2 + \color{blue}{\left(\frac{1}{2} \cdot 4\right) \cdot \left(1 - u\right)}}{v}\right) \]

   12. metadata-eval N/A

       \[\leadsto 1 + \mathsf{fma}\left(1 - u, -2, \frac{{\left(1 - u\right)}^{2} \cdot -2 + \color{blue}{2} \cdot \left(1 - u\right)}{v}\right) \]

   13. lower-fma.f32 N/A

       \[\leadsto 1 + \mathsf{fma}\left(1 - u, -2, \frac{\color{blue}{\mathsf{fma}\left({\left(1 - u\right)}^{2}, -2, 2 \cdot \left(1 - u\right)\right)}}{v}\right) \]

   14. lower-pow.f32 N/A

       \[\leadsto 1 + \mathsf{fma}\left(1 - u, -2, \frac{\mathsf{fma}\left(\color{blue}{{\left(1 - u\right)}^{2}}, -2, 2 \cdot \left(1 - u\right)\right)}{v}\right) \]

   15. lower--.f32 N/A

       \[\leadsto 1 + \mathsf{fma}\left(1 - u, -2, \frac{\mathsf{fma}\left({\color{blue}{\left(1 - u\right)}}^{2}, -2, 2 \cdot \left(1 - u\right)\right)}{v}\right) \]

   16. *-commutative N/A

       \[\leadsto 1 + \mathsf{fma}\left(1 - u, -2, \frac{\mathsf{fma}\left({\left(1 - u\right)}^{2}, -2, \color{blue}{\left(1 - u\right) \cdot 2}\right)}{v}\right) \]

   17. lower-*.f32 N/A

       \[\leadsto 1 + \mathsf{fma}\left(1 - u, -2, \frac{\mathsf{fma}\left({\left(1 - u\right)}^{2}, -2, \color{blue}{\left(1 - u\right) \cdot 2}\right)}{v}\right) \]

   18. lower--.f32 13.4

       \[\leadsto 1 + \mathsf{fma}\left(1 - u, -2, \frac{\mathsf{fma}\left({\left(1 - u\right)}^{2}, -2, \color{blue}{\left(1 - u\right)} \cdot 2\right)}{v}\right) \]

5. Applied rewrites 13.4%

   \[\leadsto 1 + \color{blue}{\mathsf{fma}\left(1 - u, -2, \frac{\mathsf{fma}\left({\left(1 - u\right)}^{2}, -2, \left(1 - u\right) \cdot 2\right)}{v}\right)} \]

6. Taylor expanded in u around 0

   \[\leadsto 1 + \left(u \cdot \left(2 + \left(-2 \cdot \frac{u}{v} + 2 \cdot \frac{1}{v}\right)\right) - \color{blue}{2}\right) \]
7. Step-by-step derivation

8. Applied rewrites 13.4%

   \[\leadsto 1 + \mathsf{fma}\left(\frac{\mathsf{fma}\left(-2, u, 2\right)}{v} - -2, \color{blue}{u}, -2\right) \]
9. Step-by-step derivation

10. Applied rewrites 13.4%

    \[\leadsto 1 + \mathsf{fma}\left(\frac{\mathsf{fma}\left(-2, u, 1\right) + 1}{v} - -2, u, -2\right) \]
11. Add Preprocessing

Alternative 12: 14.2% accurate, 7.7× speedup

Math

\[\begin{array}{l} \\ 1 + \mathsf{fma}\left(\frac{\mathsf{fma}\left(-2, u, 2\right)}{v} - -2, u, -2\right) \end{array} \]

FPCore

(FPCore (u v)
 :precision binary32
 (+ 1.0 (fma (- (/ (fma -2.0 u 2.0) v) -2.0) u -2.0)))

C

float code(float u, float v) {
	return 1.0f + fmaf(((fmaf(-2.0f, u, 2.0f) / v) - -2.0f), u, -2.0f);
}

Julia

function code(u, v)
	return Float32(Float32(1.0) + fma(Float32(Float32(fma(Float32(-2.0), u, Float32(2.0)) / v) - Float32(-2.0)), u, Float32(-2.0)))
end

Derivation

1. Initial program 99.6%

   \[1 + v \cdot \log \left(u + \left(1 - u\right) \cdot e^{\frac{-2}{v}}\right) \]
2. Add Preprocessing

3. Taylor expanded in v around inf

   \[\leadsto 1 + \color{blue}{\left(-2 \cdot \left(1 - u\right) + \frac{1}{2} \cdot \frac{-4 \cdot {\left(1 - u\right)}^{2} + 4 \cdot \left(1 - u\right)}{v}\right)} \]
4. Step-by-step derivation

   1. *-commutative N/A

      \[\leadsto 1 + \left(\color{blue}{\left(1 - u\right) \cdot -2} + \frac{1}{2} \cdot \frac{-4 \cdot {\left(1 - u\right)}^{2} + 4 \cdot \left(1 - u\right)}{v}\right) \]

   2. lower-fma.f32 N/A

      \[\leadsto 1 + \color{blue}{\mathsf{fma}\left(1 - u, -2, \frac{1}{2} \cdot \frac{-4 \cdot {\left(1 - u\right)}^{2} + 4 \cdot \left(1 - u\right)}{v}\right)} \]

   3. lower--.f32 N/A

      \[\leadsto 1 + \mathsf{fma}\left(\color{blue}{1 - u}, -2, \frac{1}{2} \cdot \frac{-4 \cdot {\left(1 - u\right)}^{2} + 4 \cdot \left(1 - u\right)}{v}\right) \]

   4. associate-*r/ N/A

      \[\leadsto 1 + \mathsf{fma}\left(1 - u, -2, \color{blue}{\frac{\frac{1}{2} \cdot \left(-4 \cdot {\left(1 - u\right)}^{2} + 4 \cdot \left(1 - u\right)\right)}{v}}\right) \]

   5. lower-/.f32 N/A

      \[\leadsto 1 + \mathsf{fma}\left(1 - u, -2, \color{blue}{\frac{\frac{1}{2} \cdot \left(-4 \cdot {\left(1 - u\right)}^{2} + 4 \cdot \left(1 - u\right)\right)}{v}}\right) \]

   6. distribute-lft-in N/A

      \[\leadsto 1 + \mathsf{fma}\left(1 - u, -2, \frac{\color{blue}{\frac{1}{2} \cdot \left(-4 \cdot {\left(1 - u\right)}^{2}\right) + \frac{1}{2} \cdot \left(4 \cdot \left(1 - u\right)\right)}}{v}\right) \]

   7. *-commutative N/A

      \[\leadsto 1 + \mathsf{fma}\left(1 - u, -2, \frac{\color{blue}{\left(-4 \cdot {\left(1 - u\right)}^{2}\right) \cdot \frac{1}{2}} + \frac{1}{2} \cdot \left(4 \cdot \left(1 - u\right)\right)}{v}\right) \]

   8. *-commutative N/A

      \[\leadsto 1 + \mathsf{fma}\left(1 - u, -2, \frac{\color{blue}{\left({\left(1 - u\right)}^{2} \cdot -4\right)} \cdot \frac{1}{2} + \frac{1}{2} \cdot \left(4 \cdot \left(1 - u\right)\right)}{v}\right) \]

   9. associate-*l* N/A

      \[\leadsto 1 + \mathsf{fma}\left(1 - u, -2, \frac{\color{blue}{{\left(1 - u\right)}^{2} \cdot \left(-4 \cdot \frac{1}{2}\right)} + \frac{1}{2} \cdot \left(4 \cdot \left(1 - u\right)\right)}{v}\right) \]

   10. metadata-eval N/A

       \[\leadsto 1 + \mathsf{fma}\left(1 - u, -2, \frac{{\left(1 - u\right)}^{2} \cdot \color{blue}{-2} + \frac{1}{2} \cdot \left(4 \cdot \left(1 - u\right)\right)}{v}\right) \]

   11. associate-*r* N/A

       \[\leadsto 1 + \mathsf{fma}\left(1 - u, -2, \frac{{\left(1 - u\right)}^{2} \cdot -2 + \color{blue}{\left(\frac{1}{2} \cdot 4\right) \cdot \left(1 - u\right)}}{v}\right) \]

   12. metadata-eval N/A

       \[\leadsto 1 + \mathsf{fma}\left(1 - u, -2, \frac{{\left(1 - u\right)}^{2} \cdot -2 + \color{blue}{2} \cdot \left(1 - u\right)}{v}\right) \]

   13. lower-fma.f32 N/A

       \[\leadsto 1 + \mathsf{fma}\left(1 - u, -2, \frac{\color{blue}{\mathsf{fma}\left({\left(1 - u\right)}^{2}, -2, 2 \cdot \left(1 - u\right)\right)}}{v}\right) \]

   14. lower-pow.f32 N/A

       \[\leadsto 1 + \mathsf{fma}\left(1 - u, -2, \frac{\mathsf{fma}\left(\color{blue}{{\left(1 - u\right)}^{2}}, -2, 2 \cdot \left(1 - u\right)\right)}{v}\right) \]

   15. lower--.f32 N/A

       \[\leadsto 1 + \mathsf{fma}\left(1 - u, -2, \frac{\mathsf{fma}\left({\color{blue}{\left(1 - u\right)}}^{2}, -2, 2 \cdot \left(1 - u\right)\right)}{v}\right) \]

   16. *-commutative N/A

       \[\leadsto 1 + \mathsf{fma}\left(1 - u, -2, \frac{\mathsf{fma}\left({\left(1 - u\right)}^{2}, -2, \color{blue}{\left(1 - u\right) \cdot 2}\right)}{v}\right) \]

   17. lower-*.f32 N/A

       \[\leadsto 1 + \mathsf{fma}\left(1 - u, -2, \frac{\mathsf{fma}\left({\left(1 - u\right)}^{2}, -2, \color{blue}{\left(1 - u\right) \cdot 2}\right)}{v}\right) \]

   18. lower--.f32 13.4

       \[\leadsto 1 + \mathsf{fma}\left(1 - u, -2, \frac{\mathsf{fma}\left({\left(1 - u\right)}^{2}, -2, \color{blue}{\left(1 - u\right)} \cdot 2\right)}{v}\right) \]

5. Applied rewrites 13.4%

   \[\leadsto 1 + \color{blue}{\mathsf{fma}\left(1 - u, -2, \frac{\mathsf{fma}\left({\left(1 - u\right)}^{2}, -2, \left(1 - u\right) \cdot 2\right)}{v}\right)} \]

6. Taylor expanded in u around 0

   \[\leadsto 1 + \left(u \cdot \left(2 + \left(-2 \cdot \frac{u}{v} + 2 \cdot \frac{1}{v}\right)\right) - \color{blue}{2}\right) \]
7. Step-by-step derivation

8. Applied rewrites 13.4%

   \[\leadsto 1 + \mathsf{fma}\left(\frac{\mathsf{fma}\left(-2, u, 2\right)}{v} - -2, \color{blue}{u}, -2\right) \]
9. Add Preprocessing

Alternative 13: 14.0% accurate, 10.0× speedup

Math

\[\begin{array}{l} \\ 2 \cdot \left(u + \frac{u}{v}\right) - 1 \end{array} \]

FPCore

(FPCore (u v) :precision binary32 (- (* 2.0 (+ u (/ u v))) 1.0))

C

float code(float u, float v) {
	return (2.0f * (u + (u / v))) - 1.0f;
}

Fortran

module fmin_fmax_functions
    implicit none
    private
    public fmax
    public fmin

    interface fmax
        module procedure fmax88
        module procedure fmax44
        module procedure fmax84
        module procedure fmax48
    end interface
    interface fmin
        module procedure fmin88
        module procedure fmin44
        module procedure fmin84
        module procedure fmin48
    end interface
contains
    real(8) function fmax88(x, y) result (res)
        real(8), intent (in) :: x
        real(8), intent (in) :: y
        res = merge(y, merge(x, max(x, y), y /= y), x /= x)
    end function
    real(4) function fmax44(x, y) result (res)
        real(4), intent (in) :: x
        real(4), intent (in) :: y
        res = merge(y, merge(x, max(x, y), y /= y), x /= x)
    end function
    real(8) function fmax84(x, y) result(res)
        real(8), intent (in) :: x
        real(4), intent (in) :: y
        res = merge(dble(y), merge(x, max(x, dble(y)), y /= y), x /= x)
    end function
    real(8) function fmax48(x, y) result(res)
        real(4), intent (in) :: x
        real(8), intent (in) :: y
        res = merge(y, merge(dble(x), max(dble(x), y), y /= y), x /= x)
    end function
    real(8) function fmin88(x, y) result (res)
        real(8), intent (in) :: x
        real(8), intent (in) :: y
        res = merge(y, merge(x, min(x, y), y /= y), x /= x)
    end function
    real(4) function fmin44(x, y) result (res)
        real(4), intent (in) :: x
        real(4), intent (in) :: y
        res = merge(y, merge(x, min(x, y), y /= y), x /= x)
    end function
    real(8) function fmin84(x, y) result(res)
        real(8), intent (in) :: x
        real(4), intent (in) :: y
        res = merge(dble(y), merge(x, min(x, dble(y)), y /= y), x /= x)
    end function
    real(8) function fmin48(x, y) result(res)
        real(4), intent (in) :: x
        real(8), intent (in) :: y
        res = merge(y, merge(dble(x), min(dble(x), y), y /= y), x /= x)
    end function
end module

real(4) function code(u, v)
use fmin_fmax_functions
    real(4), intent (in) :: u
    real(4), intent (in) :: v
    code = (2.0e0 * (u + (u / v))) - 1.0e0
end function

Julia

function code(u, v)
	return Float32(Float32(Float32(2.0) * Float32(u + Float32(u / v))) - Float32(1.0))
end

MATLAB

function tmp = code(u, v)
	tmp = (single(2.0) * (u + (u / v))) - single(1.0);
end

Derivation

1. Initial program 99.6%

   \[1 + v \cdot \log \left(u + \left(1 - u\right) \cdot e^{\frac{-2}{v}}\right) \]
2. Add Preprocessing

3. Taylor expanded in u around 0

   \[\leadsto \color{blue}{u \cdot \left(v \cdot \left(\frac{1}{e^{\frac{-2}{v}}} - 1\right)\right) - 1} \]
4. Step-by-step derivation

   1. metadata-eval N/A

      \[\leadsto u \cdot \left(v \cdot \left(\frac{1}{e^{\frac{-2}{v}}} - 1\right)\right) - \color{blue}{1 \cdot 1} \]

   2. fp-cancel-sub-sign-inv N/A

      \[\leadsto \color{blue}{u \cdot \left(v \cdot \left(\frac{1}{e^{\frac{-2}{v}}} - 1\right)\right) + \left(\mathsf{neg}\left(1\right)\right) \cdot 1} \]

   3. associate-*r* N/A

      \[\leadsto \color{blue}{\left(u \cdot v\right) \cdot \left(\frac{1}{e^{\frac{-2}{v}}} - 1\right)} + \left(\mathsf{neg}\left(1\right)\right) \cdot 1 \]

   4. metadata-eval N/A

      \[\leadsto \left(u \cdot v\right) \cdot \left(\frac{1}{e^{\frac{-2}{v}}} - 1\right) + \color{blue}{-1} \cdot 1 \]

   5. metadata-eval N/A

      \[\leadsto \left(u \cdot v\right) \cdot \left(\frac{1}{e^{\frac{-2}{v}}} - 1\right) + \color{blue}{-1} \]

   6. lower-fma.f32 N/A

      \[\leadsto \color{blue}{\mathsf{fma}\left(u \cdot v, \frac{1}{e^{\frac{-2}{v}}} - 1, -1\right)} \]

   7. lower-*.f32 N/A

      \[\leadsto \mathsf{fma}\left(\color{blue}{u \cdot v}, \frac{1}{e^{\frac{-2}{v}}} - 1, -1\right) \]

   8. rec-exp N/A

      \[\leadsto \mathsf{fma}\left(u \cdot v, \color{blue}{e^{\mathsf{neg}\left(\frac{-2}{v}\right)}} - 1, -1\right) \]

   9. distribute-frac-neg N/A

      \[\leadsto \mathsf{fma}\left(u \cdot v, e^{\color{blue}{\frac{\mathsf{neg}\left(-2\right)}{v}}} - 1, -1\right) \]

   10. metadata-eval N/A

       \[\leadsto \mathsf{fma}\left(u \cdot v, e^{\frac{\color{blue}{2}}{v}} - 1, -1\right) \]

   11. metadata-eval N/A

       \[\leadsto \mathsf{fma}\left(u \cdot v, e^{\frac{\color{blue}{2 \cdot 1}}{v}} - 1, -1\right) \]

   12. associate-*r/ N/A

       \[\leadsto \mathsf{fma}\left(u \cdot v, e^{\color{blue}{2 \cdot \frac{1}{v}}} - 1, -1\right) \]

   13. lower-expm1.f32 N/A

       \[\leadsto \mathsf{fma}\left(u \cdot v, \color{blue}{\mathsf{expm1}\left(2 \cdot \frac{1}{v}\right)}, -1\right) \]

   14. associate-*r/ N/A

       \[\leadsto \mathsf{fma}\left(u \cdot v, \mathsf{expm1}\left(\color{blue}{\frac{2 \cdot 1}{v}}\right), -1\right) \]

   15. metadata-eval N/A

       \[\leadsto \mathsf{fma}\left(u \cdot v, \mathsf{expm1}\left(\frac{\color{blue}{2}}{v}\right), -1\right) \]

   16. lower-/.f32 9.4

       \[\leadsto \mathsf{fma}\left(u \cdot v, \mathsf{expm1}\left(\color{blue}{\frac{2}{v}}\right), -1\right) \]

5. Applied rewrites 9.4%

   \[\leadsto \color{blue}{\mathsf{fma}\left(u \cdot v, \mathsf{expm1}\left(\frac{2}{v}\right), -1\right)} \]

6. Taylor expanded in v around inf

   \[\leadsto \left(2 \cdot u + 2 \cdot \frac{u}{v}\right) - \color{blue}{1} \]
7. Step-by-step derivation

8. Applied rewrites 13.3%

   \[\leadsto 2 \cdot \left(u + \frac{u}{v}\right) - \color{blue}{1} \]
9. Add Preprocessing

Alternative 14: 8.5% accurate, 10.5× speedup

Math

\[\begin{array}{l} \\ \frac{u}{v \cdot v} \cdot 1.3333333333333333 \end{array} \]

FPCore

(FPCore (u v) :precision binary32 (* (/ u (* v v)) 1.3333333333333333))

C

float code(float u, float v) {
	return (u / (v * v)) * 1.3333333333333333f;
}

Fortran

module fmin_fmax_functions
    implicit none
    private
    public fmax
    public fmin

    interface fmax
        module procedure fmax88
        module procedure fmax44
        module procedure fmax84
        module procedure fmax48
    end interface
    interface fmin
        module procedure fmin88
        module procedure fmin44
        module procedure fmin84
        module procedure fmin48
    end interface
contains
    real(8) function fmax88(x, y) result (res)
        real(8), intent (in) :: x
        real(8), intent (in) :: y
        res = merge(y, merge(x, max(x, y), y /= y), x /= x)
    end function
    real(4) function fmax44(x, y) result (res)
        real(4), intent (in) :: x
        real(4), intent (in) :: y
        res = merge(y, merge(x, max(x, y), y /= y), x /= x)
    end function
    real(8) function fmax84(x, y) result(res)
        real(8), intent (in) :: x
        real(4), intent (in) :: y
        res = merge(dble(y), merge(x, max(x, dble(y)), y /= y), x /= x)
    end function
    real(8) function fmax48(x, y) result(res)
        real(4), intent (in) :: x
        real(8), intent (in) :: y
        res = merge(y, merge(dble(x), max(dble(x), y), y /= y), x /= x)
    end function
    real(8) function fmin88(x, y) result (res)
        real(8), intent (in) :: x
        real(8), intent (in) :: y
        res = merge(y, merge(x, min(x, y), y /= y), x /= x)
    end function
    real(4) function fmin44(x, y) result (res)
        real(4), intent (in) :: x
        real(4), intent (in) :: y
        res = merge(y, merge(x, min(x, y), y /= y), x /= x)
    end function
    real(8) function fmin84(x, y) result(res)
        real(8), intent (in) :: x
        real(4), intent (in) :: y
        res = merge(dble(y), merge(x, min(x, dble(y)), y /= y), x /= x)
    end function
    real(8) function fmin48(x, y) result(res)
        real(4), intent (in) :: x
        real(8), intent (in) :: y
        res = merge(y, merge(dble(x), min(dble(x), y), y /= y), x /= x)
    end function
end module

real(4) function code(u, v)
use fmin_fmax_functions
    real(4), intent (in) :: u
    real(4), intent (in) :: v
    code = (u / (v * v)) * 1.3333333333333333e0
end function

Julia

function code(u, v)
	return Float32(Float32(u / Float32(v * v)) * Float32(1.3333333333333333))
end

MATLAB

function tmp = code(u, v)
	tmp = (u / (v * v)) * single(1.3333333333333333);
end

Derivation

1. Initial program 99.6%

   \[1 + v \cdot \log \left(u + \left(1 - u\right) \cdot e^{\frac{-2}{v}}\right) \]
2. Add Preprocessing

3. Taylor expanded in u around 0

   \[\leadsto \color{blue}{u \cdot \left(v \cdot \left(\frac{1}{e^{\frac{-2}{v}}} - 1\right)\right) - 1} \]
4. Step-by-step derivation

   1. metadata-eval N/A

      \[\leadsto u \cdot \left(v \cdot \left(\frac{1}{e^{\frac{-2}{v}}} - 1\right)\right) - \color{blue}{1 \cdot 1} \]

   2. fp-cancel-sub-sign-inv N/A

      \[\leadsto \color{blue}{u \cdot \left(v \cdot \left(\frac{1}{e^{\frac{-2}{v}}} - 1\right)\right) + \left(\mathsf{neg}\left(1\right)\right) \cdot 1} \]

   3. associate-*r* N/A

      \[\leadsto \color{blue}{\left(u \cdot v\right) \cdot \left(\frac{1}{e^{\frac{-2}{v}}} - 1\right)} + \left(\mathsf{neg}\left(1\right)\right) \cdot 1 \]

   4. metadata-eval N/A

      \[\leadsto \left(u \cdot v\right) \cdot \left(\frac{1}{e^{\frac{-2}{v}}} - 1\right) + \color{blue}{-1} \cdot 1 \]

   5. metadata-eval N/A

      \[\leadsto \left(u \cdot v\right) \cdot \left(\frac{1}{e^{\frac{-2}{v}}} - 1\right) + \color{blue}{-1} \]

   6. lower-fma.f32 N/A

      \[\leadsto \color{blue}{\mathsf{fma}\left(u \cdot v, \frac{1}{e^{\frac{-2}{v}}} - 1, -1\right)} \]

   7. lower-*.f32 N/A

      \[\leadsto \mathsf{fma}\left(\color{blue}{u \cdot v}, \frac{1}{e^{\frac{-2}{v}}} - 1, -1\right) \]

   8. rec-exp N/A

      \[\leadsto \mathsf{fma}\left(u \cdot v, \color{blue}{e^{\mathsf{neg}\left(\frac{-2}{v}\right)}} - 1, -1\right) \]

   9. distribute-frac-neg N/A

      \[\leadsto \mathsf{fma}\left(u \cdot v, e^{\color{blue}{\frac{\mathsf{neg}\left(-2\right)}{v}}} - 1, -1\right) \]

   10. metadata-eval N/A

       \[\leadsto \mathsf{fma}\left(u \cdot v, e^{\frac{\color{blue}{2}}{v}} - 1, -1\right) \]

   11. metadata-eval N/A

       \[\leadsto \mathsf{fma}\left(u \cdot v, e^{\frac{\color{blue}{2 \cdot 1}}{v}} - 1, -1\right) \]

   12. associate-*r/ N/A

       \[\leadsto \mathsf{fma}\left(u \cdot v, e^{\color{blue}{2 \cdot \frac{1}{v}}} - 1, -1\right) \]

   13. lower-expm1.f32 N/A

       \[\leadsto \mathsf{fma}\left(u \cdot v, \color{blue}{\mathsf{expm1}\left(2 \cdot \frac{1}{v}\right)}, -1\right) \]

   14. associate-*r/ N/A

       \[\leadsto \mathsf{fma}\left(u \cdot v, \mathsf{expm1}\left(\color{blue}{\frac{2 \cdot 1}{v}}\right), -1\right) \]

   15. metadata-eval N/A

       \[\leadsto \mathsf{fma}\left(u \cdot v, \mathsf{expm1}\left(\frac{\color{blue}{2}}{v}\right), -1\right) \]

   16. lower-/.f32 9.4

       \[\leadsto \mathsf{fma}\left(u \cdot v, \mathsf{expm1}\left(\color{blue}{\frac{2}{v}}\right), -1\right) \]

5. Applied rewrites 9.4%

   \[\leadsto \color{blue}{\mathsf{fma}\left(u \cdot v, \mathsf{expm1}\left(\frac{2}{v}\right), -1\right)} \]

6. Taylor expanded in v around inf

   \[\leadsto \left(\frac{4}{3} \cdot \frac{u}{{v}^{2}} + \left(2 \cdot u + 2 \cdot \frac{u}{v}\right)\right) - \color{blue}{1} \]
7. Step-by-step derivation

8. Applied rewrites 11.5%

   \[\leadsto \mathsf{fma}\left(\frac{1.3333333333333333}{v}, \frac{u}{v}, 2 \cdot \left(u + \frac{u}{v}\right)\right) - \color{blue}{1} \]

9. Taylor expanded in v around 0

   \[\leadsto \frac{4}{3} \cdot \frac{u}{\color{blue}{{v}^{2}}} \]
10. Step-by-step derivation

11. Applied rewrites 9.1%

    \[\leadsto \frac{u}{v \cdot v} \cdot 1.3333333333333333 \]
12. Add Preprocessing

Alternative 15: 8.0% accurate, 11.6× speedup

Math

\[\begin{array}{l} \\ \left(2 - \frac{1}{u}\right) \cdot u \end{array} \]

FPCore

(FPCore (u v) :precision binary32 (* (- 2.0 (/ 1.0 u)) u))

C

float code(float u, float v) {
	return (2.0f - (1.0f / u)) * u;
}

Fortran

module fmin_fmax_functions
    implicit none
    private
    public fmax
    public fmin

    interface fmax
        module procedure fmax88
        module procedure fmax44
        module procedure fmax84
        module procedure fmax48
    end interface
    interface fmin
        module procedure fmin88
        module procedure fmin44
        module procedure fmin84
        module procedure fmin48
    end interface
contains
    real(8) function fmax88(x, y) result (res)
        real(8), intent (in) :: x
        real(8), intent (in) :: y
        res = merge(y, merge(x, max(x, y), y /= y), x /= x)
    end function
    real(4) function fmax44(x, y) result (res)
        real(4), intent (in) :: x
        real(4), intent (in) :: y
        res = merge(y, merge(x, max(x, y), y /= y), x /= x)
    end function
    real(8) function fmax84(x, y) result(res)
        real(8), intent (in) :: x
        real(4), intent (in) :: y
        res = merge(dble(y), merge(x, max(x, dble(y)), y /= y), x /= x)
    end function
    real(8) function fmax48(x, y) result(res)
        real(4), intent (in) :: x
        real(8), intent (in) :: y
        res = merge(y, merge(dble(x), max(dble(x), y), y /= y), x /= x)
    end function
    real(8) function fmin88(x, y) result (res)
        real(8), intent (in) :: x
        real(8), intent (in) :: y
        res = merge(y, merge(x, min(x, y), y /= y), x /= x)
    end function
    real(4) function fmin44(x, y) result (res)
        real(4), intent (in) :: x
        real(4), intent (in) :: y
        res = merge(y, merge(x, min(x, y), y /= y), x /= x)
    end function
    real(8) function fmin84(x, y) result(res)
        real(8), intent (in) :: x
        real(4), intent (in) :: y
        res = merge(dble(y), merge(x, min(x, dble(y)), y /= y), x /= x)
    end function
    real(8) function fmin48(x, y) result(res)
        real(4), intent (in) :: x
        real(8), intent (in) :: y
        res = merge(y, merge(dble(x), min(dble(x), y), y /= y), x /= x)
    end function
end module

real(4) function code(u, v)
use fmin_fmax_functions
    real(4), intent (in) :: u
    real(4), intent (in) :: v
    code = (2.0e0 - (1.0e0 / u)) * u
end function

Julia

function code(u, v)
	return Float32(Float32(Float32(2.0) - Float32(Float32(1.0) / u)) * u)
end

MATLAB

function tmp = code(u, v)
	tmp = (single(2.0) - (single(1.0) / u)) * u;
end

Derivation

1. Initial program 99.6%

   \[1 + v \cdot \log \left(u + \left(1 - u\right) \cdot e^{\frac{-2}{v}}\right) \]
2. Add Preprocessing

3. Taylor expanded in u around 0

   \[\leadsto \color{blue}{u \cdot \left(v \cdot \left(\frac{1}{e^{\frac{-2}{v}}} - 1\right)\right) - 1} \]
4. Step-by-step derivation

   1. metadata-eval N/A

      \[\leadsto u \cdot \left(v \cdot \left(\frac{1}{e^{\frac{-2}{v}}} - 1\right)\right) - \color{blue}{1 \cdot 1} \]

   2. fp-cancel-sub-sign-inv N/A

      \[\leadsto \color{blue}{u \cdot \left(v \cdot \left(\frac{1}{e^{\frac{-2}{v}}} - 1\right)\right) + \left(\mathsf{neg}\left(1\right)\right) \cdot 1} \]

   3. associate-*r* N/A

      \[\leadsto \color{blue}{\left(u \cdot v\right) \cdot \left(\frac{1}{e^{\frac{-2}{v}}} - 1\right)} + \left(\mathsf{neg}\left(1\right)\right) \cdot 1 \]

   4. metadata-eval N/A

      \[\leadsto \left(u \cdot v\right) \cdot \left(\frac{1}{e^{\frac{-2}{v}}} - 1\right) + \color{blue}{-1} \cdot 1 \]

   5. metadata-eval N/A

      \[\leadsto \left(u \cdot v\right) \cdot \left(\frac{1}{e^{\frac{-2}{v}}} - 1\right) + \color{blue}{-1} \]

   6. lower-fma.f32 N/A

      \[\leadsto \color{blue}{\mathsf{fma}\left(u \cdot v, \frac{1}{e^{\frac{-2}{v}}} - 1, -1\right)} \]

   7. lower-*.f32 N/A

      \[\leadsto \mathsf{fma}\left(\color{blue}{u \cdot v}, \frac{1}{e^{\frac{-2}{v}}} - 1, -1\right) \]

   8. rec-exp N/A

      \[\leadsto \mathsf{fma}\left(u \cdot v, \color{blue}{e^{\mathsf{neg}\left(\frac{-2}{v}\right)}} - 1, -1\right) \]

   9. distribute-frac-neg N/A

      \[\leadsto \mathsf{fma}\left(u \cdot v, e^{\color{blue}{\frac{\mathsf{neg}\left(-2\right)}{v}}} - 1, -1\right) \]

   10. metadata-eval N/A

       \[\leadsto \mathsf{fma}\left(u \cdot v, e^{\frac{\color{blue}{2}}{v}} - 1, -1\right) \]

   11. metadata-eval N/A

       \[\leadsto \mathsf{fma}\left(u \cdot v, e^{\frac{\color{blue}{2 \cdot 1}}{v}} - 1, -1\right) \]

   12. associate-*r/ N/A

       \[\leadsto \mathsf{fma}\left(u \cdot v, e^{\color{blue}{2 \cdot \frac{1}{v}}} - 1, -1\right) \]

   13. lower-expm1.f32 N/A

       \[\leadsto \mathsf{fma}\left(u \cdot v, \color{blue}{\mathsf{expm1}\left(2 \cdot \frac{1}{v}\right)}, -1\right) \]

   14. associate-*r/ N/A

       \[\leadsto \mathsf{fma}\left(u \cdot v, \mathsf{expm1}\left(\color{blue}{\frac{2 \cdot 1}{v}}\right), -1\right) \]

   15. metadata-eval N/A

       \[\leadsto \mathsf{fma}\left(u \cdot v, \mathsf{expm1}\left(\frac{\color{blue}{2}}{v}\right), -1\right) \]

   16. lower-/.f32 9.4

       \[\leadsto \mathsf{fma}\left(u \cdot v, \mathsf{expm1}\left(\color{blue}{\frac{2}{v}}\right), -1\right) \]

5. Applied rewrites 9.4%

   \[\leadsto \color{blue}{\mathsf{fma}\left(u \cdot v, \mathsf{expm1}\left(\frac{2}{v}\right), -1\right)} \]

6. Taylor expanded in v around inf

   \[\leadsto 2 \cdot u - \color{blue}{1} \]
7. Step-by-step derivation

8. Applied rewrites 7.4%

   \[\leadsto 2 \cdot u - \color{blue}{1} \]

9. Taylor expanded in u around inf

   \[\leadsto u \cdot \left(2 - \color{blue}{\frac{1}{u}}\right) \]
10. Step-by-step derivation

11. Applied rewrites 7.4%

    \[\leadsto \left(2 - \frac{1}{u}\right) \cdot u \]
12. Add Preprocessing

Alternative 16: 8.0% accurate, 33.0× speedup

Math

\[\begin{array}{l} \\ u + \left(u - 1\right) \end{array} \]

FPCore

(FPCore (u v) :precision binary32 (+ u (- u 1.0)))

C

float code(float u, float v) {
	return u + (u - 1.0f);
}

Fortran

module fmin_fmax_functions
    implicit none
    private
    public fmax
    public fmin

    interface fmax
        module procedure fmax88
        module procedure fmax44
        module procedure fmax84
        module procedure fmax48
    end interface
    interface fmin
        module procedure fmin88
        module procedure fmin44
        module procedure fmin84
        module procedure fmin48
    end interface
contains
    real(8) function fmax88(x, y) result (res)
        real(8), intent (in) :: x
        real(8), intent (in) :: y
        res = merge(y, merge(x, max(x, y), y /= y), x /= x)
    end function
    real(4) function fmax44(x, y) result (res)
        real(4), intent (in) :: x
        real(4), intent (in) :: y
        res = merge(y, merge(x, max(x, y), y /= y), x /= x)
    end function
    real(8) function fmax84(x, y) result(res)
        real(8), intent (in) :: x
        real(4), intent (in) :: y
        res = merge(dble(y), merge(x, max(x, dble(y)), y /= y), x /= x)
    end function
    real(8) function fmax48(x, y) result(res)
        real(4), intent (in) :: x
        real(8), intent (in) :: y
        res = merge(y, merge(dble(x), max(dble(x), y), y /= y), x /= x)
    end function
    real(8) function fmin88(x, y) result (res)
        real(8), intent (in) :: x
        real(8), intent (in) :: y
        res = merge(y, merge(x, min(x, y), y /= y), x /= x)
    end function
    real(4) function fmin44(x, y) result (res)
        real(4), intent (in) :: x
        real(4), intent (in) :: y
        res = merge(y, merge(x, min(x, y), y /= y), x /= x)
    end function
    real(8) function fmin84(x, y) result(res)
        real(8), intent (in) :: x
        real(4), intent (in) :: y
        res = merge(dble(y), merge(x, min(x, dble(y)), y /= y), x /= x)
    end function
    real(8) function fmin48(x, y) result(res)
        real(4), intent (in) :: x
        real(8), intent (in) :: y
        res = merge(y, merge(dble(x), min(dble(x), y), y /= y), x /= x)
    end function
end module

real(4) function code(u, v)
use fmin_fmax_functions
    real(4), intent (in) :: u
    real(4), intent (in) :: v
    code = u + (u - 1.0e0)
end function

Julia

function code(u, v)
	return Float32(u + Float32(u - Float32(1.0)))
end

MATLAB

function tmp = code(u, v)
	tmp = u + (u - single(1.0));
end

Derivation

1. Initial program 99.6%

   \[1 + v \cdot \log \left(u + \left(1 - u\right) \cdot e^{\frac{-2}{v}}\right) \]
2. Add Preprocessing

3. Taylor expanded in u around 0

   \[\leadsto \color{blue}{u \cdot \left(v \cdot \left(\frac{1}{e^{\frac{-2}{v}}} - 1\right)\right) - 1} \]
4. Step-by-step derivation

   1. metadata-eval N/A

      \[\leadsto u \cdot \left(v \cdot \left(\frac{1}{e^{\frac{-2}{v}}} - 1\right)\right) - \color{blue}{1 \cdot 1} \]

   2. fp-cancel-sub-sign-inv N/A

      \[\leadsto \color{blue}{u \cdot \left(v \cdot \left(\frac{1}{e^{\frac{-2}{v}}} - 1\right)\right) + \left(\mathsf{neg}\left(1\right)\right) \cdot 1} \]

   3. associate-*r* N/A

      \[\leadsto \color{blue}{\left(u \cdot v\right) \cdot \left(\frac{1}{e^{\frac{-2}{v}}} - 1\right)} + \left(\mathsf{neg}\left(1\right)\right) \cdot 1 \]

   4. metadata-eval N/A

      \[\leadsto \left(u \cdot v\right) \cdot \left(\frac{1}{e^{\frac{-2}{v}}} - 1\right) + \color{blue}{-1} \cdot 1 \]

   5. metadata-eval N/A

      \[\leadsto \left(u \cdot v\right) \cdot \left(\frac{1}{e^{\frac{-2}{v}}} - 1\right) + \color{blue}{-1} \]

   6. lower-fma.f32 N/A

      \[\leadsto \color{blue}{\mathsf{fma}\left(u \cdot v, \frac{1}{e^{\frac{-2}{v}}} - 1, -1\right)} \]

   7. lower-*.f32 N/A

      \[\leadsto \mathsf{fma}\left(\color{blue}{u \cdot v}, \frac{1}{e^{\frac{-2}{v}}} - 1, -1\right) \]

   8. rec-exp N/A

      \[\leadsto \mathsf{fma}\left(u \cdot v, \color{blue}{e^{\mathsf{neg}\left(\frac{-2}{v}\right)}} - 1, -1\right) \]

   9. distribute-frac-neg N/A

      \[\leadsto \mathsf{fma}\left(u \cdot v, e^{\color{blue}{\frac{\mathsf{neg}\left(-2\right)}{v}}} - 1, -1\right) \]

   10. metadata-eval N/A

       \[\leadsto \mathsf{fma}\left(u \cdot v, e^{\frac{\color{blue}{2}}{v}} - 1, -1\right) \]

   11. metadata-eval N/A

       \[\leadsto \mathsf{fma}\left(u \cdot v, e^{\frac{\color{blue}{2 \cdot 1}}{v}} - 1, -1\right) \]

   12. associate-*r/ N/A

       \[\leadsto \mathsf{fma}\left(u \cdot v, e^{\color{blue}{2 \cdot \frac{1}{v}}} - 1, -1\right) \]

   13. lower-expm1.f32 N/A

       \[\leadsto \mathsf{fma}\left(u \cdot v, \color{blue}{\mathsf{expm1}\left(2 \cdot \frac{1}{v}\right)}, -1\right) \]

   14. associate-*r/ N/A

       \[\leadsto \mathsf{fma}\left(u \cdot v, \mathsf{expm1}\left(\color{blue}{\frac{2 \cdot 1}{v}}\right), -1\right) \]

   15. metadata-eval N/A

       \[\leadsto \mathsf{fma}\left(u \cdot v, \mathsf{expm1}\left(\frac{\color{blue}{2}}{v}\right), -1\right) \]

   16. lower-/.f32 9.4

       \[\leadsto \mathsf{fma}\left(u \cdot v, \mathsf{expm1}\left(\color{blue}{\frac{2}{v}}\right), -1\right) \]

5. Applied rewrites 9.4%

   \[\leadsto \color{blue}{\mathsf{fma}\left(u \cdot v, \mathsf{expm1}\left(\frac{2}{v}\right), -1\right)} \]

6. Taylor expanded in v around inf

   \[\leadsto 2 \cdot u - \color{blue}{1} \]
7. Step-by-step derivation

8. Applied rewrites 7.4%

   \[\leadsto 2 \cdot u - \color{blue}{1} \]
9. Step-by-step derivation

10. Applied rewrites 7.4%

    \[\leadsto u + \left(u - \color{blue}{1}\right) \]
11. Add Preprocessing

Alternative 17: 5.8% accurate, 231.0× speedup

Math

\[\begin{array}{l} \\ -1 \end{array} \]

FPCore

(FPCore (u v) :precision binary32 -1.0)

C

float code(float u, float v) {
	return -1.0f;
}

Fortran

module fmin_fmax_functions
    implicit none
    private
    public fmax
    public fmin

    interface fmax
        module procedure fmax88
        module procedure fmax44
        module procedure fmax84
        module procedure fmax48
    end interface
    interface fmin
        module procedure fmin88
        module procedure fmin44
        module procedure fmin84
        module procedure fmin48
    end interface
contains
    real(8) function fmax88(x, y) result (res)
        real(8), intent (in) :: x
        real(8), intent (in) :: y
        res = merge(y, merge(x, max(x, y), y /= y), x /= x)
    end function
    real(4) function fmax44(x, y) result (res)
        real(4), intent (in) :: x
        real(4), intent (in) :: y
        res = merge(y, merge(x, max(x, y), y /= y), x /= x)
    end function
    real(8) function fmax84(x, y) result(res)
        real(8), intent (in) :: x
        real(4), intent (in) :: y
        res = merge(dble(y), merge(x, max(x, dble(y)), y /= y), x /= x)
    end function
    real(8) function fmax48(x, y) result(res)
        real(4), intent (in) :: x
        real(8), intent (in) :: y
        res = merge(y, merge(dble(x), max(dble(x), y), y /= y), x /= x)
    end function
    real(8) function fmin88(x, y) result (res)
        real(8), intent (in) :: x
        real(8), intent (in) :: y
        res = merge(y, merge(x, min(x, y), y /= y), x /= x)
    end function
    real(4) function fmin44(x, y) result (res)
        real(4), intent (in) :: x
        real(4), intent (in) :: y
        res = merge(y, merge(x, min(x, y), y /= y), x /= x)
    end function
    real(8) function fmin84(x, y) result(res)
        real(8), intent (in) :: x
        real(4), intent (in) :: y
        res = merge(dble(y), merge(x, min(x, dble(y)), y /= y), x /= x)
    end function
    real(8) function fmin48(x, y) result(res)
        real(4), intent (in) :: x
        real(8), intent (in) :: y
        res = merge(y, merge(dble(x), min(dble(x), y), y /= y), x /= x)
    end function
end module

real(4) function code(u, v)
use fmin_fmax_functions
    real(4), intent (in) :: u
    real(4), intent (in) :: v
    code = -1.0e0
end function

Julia

function code(u, v)
	return Float32(-1.0)
end

MATLAB

function tmp = code(u, v)
	tmp = single(-1.0);
end

Derivation

1. Initial program 99.6%

   \[1 + v \cdot \log \left(u + \left(1 - u\right) \cdot e^{\frac{-2}{v}}\right) \]
2. Add Preprocessing

3. Taylor expanded in u around 0

   \[\leadsto \color{blue}{-1} \]
4. Step-by-step derivation

5. Applied rewrites 5.1%

   \[\leadsto \color{blue}{-1} \]
6. Add Preprocessing

Reproduce

herbie shell --seed 2025017 
(FPCore (u v)
  :name "HairBSDF, sample_f, cosTheta"
  :precision binary32
  :pre (and (and (<= 1e-5 u) (<= u 1.0)) (and (<= 0.0 v) (<= v 109.746574)))
  (+ 1.0 (* v (log (+ u (* (- 1.0 u) (exp (/ -2.0 v))))))))