kernels.variance

kernels.variance(arr, *, ddof=0, return_mean=False, validate=True)

Return the NaN-aware variance along the stack axis.

Parameters

arr : (ndarray, shape(N, *spatial))

Image stack. Accepted dtypes are uint8, uint16, int16, int32, float32, and float64. Integer inputs are promoted to the package’s floating workspace when validate is True. Inputs with more than 3 dimensions are flattened internally; output shapes match the trailing spatial dimensions of the input.

ddof : int = 0

Delta degrees of freedom. The returned value is sum((valid - mean)**2) / (nvalid - ddof). Pixels with nvalid <= ddof return NaN.

return_mean : bool = False

If True, also return the per-pixel mean from the same accumulation pass.

validate : bool = True

If True, check dimensionality and normalize dtype/contiguity before entering the Rust kernel. If False, callers must provide inputs that satisfy the compiled kernel assumptions.

Returns

variance : (ndarray, shape(*spatial))

Per-pixel variance of finite values. Use np.sqrt(var) if a standard-deviation or error-like map is needed.

mean : (ndarray, shape(*spatial))

Returned only when return_mean=True.