GpuMath

1 min read

Import: import system.gpu;

This reference lists the public declarations for this part of the library. See Reading API signatures for parameter and result notation. Shader-only declarations can be used only inside shader entry points and helpers reached from them; using them elsewhere is a compile error.

This type spans several pages: part 1 · part 2.

GpuMath

Math functions for shaders. Each works on float, float16, GpuVec2, GpuVec3 and GpuVec4, component by component, and returns the same type as its arguments. All of them except step, sign, floor, ceil, round and fract are differentiable inside @differentiable helpers. They exist only on the GPU; on the CPU use Math and the vector methods.

public static class GpuMath

sin

Sine of x radians.

public static extern fun sin<T>(x: T): T

cos

Cosine of x radians.

public static extern fun cos<T>(x: T): T

tan

Tangent of x radians.

public static extern fun tan<T>(x: T): T

asin

Arcsine in radians, for x from -1 to 1.

public static extern fun asin<T>(x: T): T

acos

Arccosine in radians, for x from -1 to 1.

public static extern fun acos<T>(x: T): T

atan

Arctangent in radians.

public static extern fun atan<T>(x: T): T

atan2

The angle in radians of the point (x, y), from -π to π.

public static extern fun atan2<T>(y: T, x: T): T

exp

e raised to x.

public static extern fun exp<T>(x: T): T

exp2

2 raised to x.

public static extern fun exp2<T>(x: T): T

log

Natural logarithm, for positive x.

public static extern fun log<T>(x: T): T

log2

Base-2 logarithm, for positive x.

public static extern fun log2<T>(x: T): T

pow

x raised to y, for non-negative x.

public static extern fun pow<T>(x: T, y: T): T

sqrt

Square root, for non-negative x.

public static extern fun sqrt<T>(x: T): T

rsqrt

1 / sqrt(x), usually faster than dividing.

public static extern fun rsqrt<T>(x: T): T

abs

Absolute value.

public static extern fun abs<T>(x: T): T

sign

-1, 0 or 1 by the sign of x.

public static extern fun sign<T>(x: T): T

floor

Rounds down to a whole number.

public static extern fun floor<T>(x: T): T

ceil

Rounds up to a whole number.

public static extern fun ceil<T>(x: T): T

round

Rounds to the nearest whole number, halves to even.

public static extern fun round<T>(x: T): T

fract

x - floor(x), from 0 up to but excluding 1.

public static extern fun fract<T>(x: T): T

min

The smaller of a and b.

public static extern fun min<T>(a: T, b: T): T

max

The larger of a and b.

public static extern fun max<T>(a: T, b: T): T

clamp

x limited to the range low through high.

public static extern fun clamp<T>(x: T, low: T, high: T): T