Utils

Some utility functions and types that are used in various places in the codebase. Ideally it should be divided into multiple very small modules.

isundef(x)

Returns true if the value x is undefined, i.e. it has not been initialized.

length_of_trimmed(predicate, x) where N

Returns the length of the collection x after removing the elements from the beginning and the end that satisfy the predicate.

⚠ The collection must have the property so that x[i] for i in firstindex(x):lastindex(x) is valid.

lerp(part, from, to)

Performs linear interpolation between from and to using part as the fraction. Equivalent to from + part * (to - from).

lerp2(dx, dy, v00, v10, v01, v11)

Performs bilinear interpolation between four values v00, v10, v01, v11 using fractions dx and dy.

lerp3(dx, dy, dz, v000, v100, v010, v110, v001, v101, v011, v111)

Performs trilinear interpolation between eight values v... using fractions dx, dy, and dz.

lerp4(a::Couple, b::Couple, c::Couple, d::Couple, dy, dx, dz)

Performs trilinear interpolation where the initial eight corner values are derived from four Couples (pairs of values) by first interpolating along the y-axis using dy. Subsequent interpolations use dz and dx.

Specifically, it computes: lerp(dx, lerp(dz, lerp(dy, a[1], a[2]), lerp(dy, c[1], c[2])), lerp(dz, lerp(dy, b[1], b[2]), lerp(dy, d[1], d[2])))

threading(kind = :off; kwargs...)

Return a Scheduler from the OhMyThreads package type that contains all the static parameters needed for threading. See the documentation of OhMyThreads related to the schedulers to know what to pass to the kwargs.

@only_float32 expr

Transforms all real literals in the expr to Float32.

Example

@only_float32 function f()
    x = 1 + 2im # expand to `1.0f0 + 2.0f0im`
    x += 1 # expand to `x += 1.0f0`
    return x
end