from_numpy

→Tensor

from_numpy(arr: np.ndarray)

source

Create a CPU tensor from a NumPy ndarray with shared storage.

The returned tensor wraps the array's existing buffer — no data is copied — and inherits the array's dtype according to the canonical NumPy → Lucid mapping (np.float32 → lucid.float32, np.int64 → lucid.int64, etc.). Because storage is shared, mutations in the array are visible in the tensor and vice versa.

Parameters

arrnumpy.ndarray

Source array. Must reside in CPU memory. Any layout (C / Fortran / strided) is accepted; the resulting tensor preserves the array's strides where possible.

Returns

Tensor

A CPU tensor sharing storage with arr.

Raises

TypeError

If arr is not a NumPy ndarray.

ValueError

If arr's dtype has no corresponding Lucid dtype (e.g. np.float128 on some platforms).

Notes

This is one of the documented H4 bridge boundaries — the only places where Lucid is allowed to take a NumPy array as input. To move the result onto a Metal device, chain Tensor.to::

t = lucid.from_numpy(arr).to("metal")

Examples

>>> import lucid, numpy as np
>>> arr = np.array([[1.0, 2.0], [3.0, 4.0]], dtype=np.float32)
>>> t = lucid.from_numpy(arr)
>>> t.dtype
float32
>>> arr[0, 0] = 99.0          # mutate the source
>>> t[0, 0].item()            # change visible in the tensor
99.0