Module cpu 

Pure-Rust CPU implementations of common raster element-wise operations.

These are used as CPU fallback kernels when the GPU is unavailable. All functions operate on &[f32] slices and return owned Vec<f32>.

When the input slices have different lengths the shorter slice determines how many elements are processed (same semantics as Iterator::zip).

Functions

abs

Absolute value: result[i] = data[i].abs().

add_scalar

Scalar addition: result[i] = data[i] + scalar.

add_slices

Element-wise addition: result[i] = a[i] + b[i].

clamp

Clamp all elements to [lo, hi]: result[i] = data[i].clamp(lo, hi).

div_scalar

Scalar division: result[i] = data[i] / scalar.

div_slices

Element-wise division: result[i] = a[i] / b[i].

max_slices

Element-wise maximum: result[i] = a[i].max(b[i]).

max_value

Maximum element value. Returns f32::MIN for empty slices.

mean

Arithmetic mean of all elements. Returns 0.0 for empty slices.

min_slices

Element-wise minimum: result[i] = a[i].min(b[i]).

min_value

Minimum element value. Returns f32::MAX for empty slices.

mul_scalar

Scalar multiplication: result[i] = data[i] * scalar.

mul_slices

Element-wise multiplication: result[i] = a[i] * b[i].

ndvi

Normalised Difference Vegetation Index: (nir - red) / (nir + red).

powf

Power: result[i] = data[i].powf(exponent).

sqrt

Square root: result[i] = data[i].sqrt().

std_dev

Population standard deviation. Returns 0.0 for slices with fewer than two elements.

sub_scalar

Scalar subtraction: result[i] = data[i] - scalar.

sub_slices

Element-wise subtraction: result[i] = a[i] - b[i].

sum

Sum of all elements. Returns 0.0 for empty slices.

variance

Population variance of all elements. Returns 0.0 for slices with fewer than two elements.