Skip to content

Fix: corrections to NaN handling for Exp and Log funcs #83

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Merged
IsaMorphic merged 1 commit into from
Mar 19, 2026
Merged

Fix: corrections to NaN handling for Exp and Log funcs #83

Changes from all commits
Commits
Show all changes
1 commit
Select commit
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
109 changes: 92 additions & 17 deletions QuadrupleLib/Modules/MathOperations.cs
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -303,7 +303,7 @@ public static Float128<TAccelerator> Acosh(Float128<TAccelerator> x)
{
return Log(x + Sqrt(FusedMultiplyAdd(x, x, NegativeOne)));
}
else if (x == One)
else if (x == One)
{
return Zero;
}
Expand Down Expand Up @@ -356,7 +356,11 @@ private static Float128<TAccelerator> _Log2(Float128<TAccelerator> y, int N)

public static Float128<TAccelerator> Log2(Float128<TAccelerator> x)
{
if (x <= Zero)
if (IsNaN(x))
{
return _qNaN;
}
else if (x <= Zero)
{
return _sNaN;
}
Expand All @@ -375,7 +379,11 @@ public static Float128<TAccelerator> Log2(Float128<TAccelerator> x)

public static Float128<TAccelerator> Log(Float128<TAccelerator> y)
{
if (y <= Zero)
if (IsNaN(y))
{
return _qNaN;
}
else if (y <= Zero)
{
return _sNaN;
}
Expand All @@ -396,7 +404,11 @@ public static Float128<TAccelerator> Log(Float128<TAccelerator> y)

public static Float128<TAccelerator> Log(Float128<TAccelerator> y, Float128<TAccelerator> newBase)
{
if (y <= Zero || newBase <= Zero)
if (IsNaN(y) || IsNaN(newBase))
{
return _qNaN;
}
else if (y <= Zero || newBase <= Zero)
{
return _sNaN;
}
Expand Down Expand Up @@ -429,7 +441,11 @@ public static Float128<TAccelerator> Log(Float128<TAccelerator> y, Float128<TAcc

public static Float128<TAccelerator> Log10(Float128<TAccelerator> y)
{
if (y <= Zero)
if (IsNaN(y))
{
return _qNaN;
}
else if (y <= Zero)
{
return _sNaN;
}
Expand All @@ -454,6 +470,11 @@ public static Float128<TAccelerator> Log10(Float128<TAccelerator> y)

public static Float128<TAccelerator> Exp(Float128<TAccelerator> y)
{
if (IsNaN(y))
{
return _qNaN;
}

Float128<TAccelerator> x_n = One;
if (y > Zero)
{
Expand All @@ -474,15 +495,20 @@ public static Float128<TAccelerator> Exp(Float128<TAccelerator> y)
{
for (int n = 0; n < 5; n++)
{
x_n = FusedMultiplyAdd(x_n, y - Log(Abs(x_n)), x_n);
x_n = Abs(FusedMultiplyAdd(x_n, y - Log(x_n), x_n));
}
}

return Abs(x_n);
return x_n;
}

public static Float128<TAccelerator> Exp10(Float128<TAccelerator> y)
{
if (IsNaN(y))
{
return _qNaN;
}

Float128<TAccelerator> x_n = One;
if (y > Zero)
{
Expand All @@ -504,32 +530,54 @@ public static Float128<TAccelerator> Exp10(Float128<TAccelerator> y)
Float128<TAccelerator> log10 = Log(10);
for (int n = 0; n < 5; n++)
{
x_n = FusedMultiplyAdd(x_n * log10, y - Log10(Abs(x_n)), x_n);
x_n = Abs(FusedMultiplyAdd(x_n * log10, y - Log10(x_n), x_n));
}
}

return Abs(x_n);
return x_n;
}

public static Float128<TAccelerator> Exp2(Float128<TAccelerator> y)
{
if (IsNaN(y))
{
return _qNaN;
}

Float128<TAccelerator> x_n = ScaleB(One, (int)y);

if (y % One != Zero)
{
Float128<TAccelerator> log2 = One / Log2(E);
for (int n = 0; n < 5; n++)
{
x_n = FusedMultiplyAdd(x_n * log2, y - Log2(Abs(x_n)), x_n);
x_n = Abs(FusedMultiplyAdd(x_n * log2, y - Log2(x_n), x_n));
}
}

return Abs(x_n);
return x_n;
}

public static Float128<TAccelerator> Pow(Float128<TAccelerator> x, Float128<TAccelerator> y)
{
return Exp(y * Log(x));
if (IsZero(x) && IsZero(y))
{
return _sNaN;
}
else if (IsInteger(x) && IsInteger(y))
{
Float128<TAccelerator> result = x;
for (int i = 1; i < y; i++)
{
result *= x;
}

return result;
}
else
{
return Exp(y * Log(x));
}
}

#endregion
Expand All @@ -538,16 +586,32 @@ public static Float128<TAccelerator> Pow(Float128<TAccelerator> x, Float128<TAcc

public static Float128<TAccelerator> Sqrt(Float128<TAccelerator> x)
{
Float128<TAccelerator> y_n = x * 0.5;
for (int n = 0; n < 25; n++)
if (IsNaN(x))
{
y_n = 0.5 * (y_n + x / y_n);
return _qNaN;
}
else if (x < Zero)
{
return _sNaN;
}
else
{
Float128<TAccelerator> y_n = x * 0.5;
for (int n = 0; n < 25; n++)
{
y_n = 0.5 * (y_n + x / y_n);
}
return y_n;
}
return y_n;
}

public static Float128<TAccelerator> Cbrt(Float128<TAccelerator> x)
{
if (IsNaN(x))
{
return _qNaN;
}

Float128<TAccelerator> y_n = x * 0.5;
for (int n = 0; n < 25; n++)
{
Expand All @@ -559,7 +623,18 @@ public static Float128<TAccelerator> Cbrt(Float128<TAccelerator> x)

public static Float128<TAccelerator> RootN(Float128<TAccelerator> x, int n)
{
return Pow(x, One / n);
if (IsNaN(x))
{
return _qNaN;
}
else if (x < Zero)
{
return (n & 1) == 0 ? _sNaN : -Pow(-x, n);
}
else
{
return Pow(x, One / n);
}
}

#endregion
Expand Down
Loading