fields.buffers.core.ArrayBuffer.as_repr

ArrayBuffer.as_repr() → _Buffer | _SupportsArray[dtype[Any]] | _NestedSequence[_SupportsArray[dtype[Any]]] | complex | bytes | str | _NestedSequence[complex | bytes | str]

Return a NumPy-compatible array for use in NumPy operations.

This method is used internally to provide a consistent interface for applying NumPy ufuncs and broadcasting logic across different buffer backends. It returns an array-like object that is suitable for NumPy operations such as np.sin(), np.add(), or reductions like np.sum().

By default, this returns the result of as_array(), which typically coerces the internal buffer into a standard numpy.ndarray. Subclasses may override this method to return more specialized representations (e.g., a unyt_array that preserves units, or a lazily sliced h5py.Dataset).

Returns:

A NumPy-operable array object, such as numpy.ndarray, unyt_array, or a backend-compatible equivalent.

Return type:

ArrayLike

See also

BufferBase.__array_ufunc__

How NumPy dispatches operations on buffers.

BufferBase.as_array

Returns the NumPy array representation used here by default.