× ⚠️ You are reading the development version documentation.
Read here the documentation of the last stable release.

Numeric backends

The core numeric path handles ordinary NumPy arrays and ordinary complex numbers. Plugins can provide additional execution branches for specialized array types without adding those dependencies to the core.

The hypercomplex plugin uses this pattern for quaternion arrays. A plugin can register:

• ndmath.execution_branch to identify a plugin-owned numeric branch;

• ndmath.execute to run an operation for that branch;

• ndmath.numpy_method.<name> to override selected NumPy-like methods;

• fft.encoding for domain-specific FFT acquisition encodings.

Example:

def register_handlers(self) -> dict:
    return {
        "ndmath.execution_branch": execution_branch,
        "ndmath.execute": execute,
    }

Handlers should return None when they do not recognize the data so the core default path remains available.