enp#

from etils import enp

Numpy API.

When possible, utils are meant to work with both numpy and jax.numpy.

All symbols#


enp	Numpy API.
enp.compat	Compat utils between TF/Torch/Numpy/Jax.
enp.dtypes	Dtype utils.
enp.linalg	`linalg` compat module for interoperability between tf, jax, numpy.
enp.typing	Typing utils.


enp.ArraySpec	Structure containing shape/dtype.
enp.NpModule	Special type indicating an unconstrained type.
enp.dtypes.AnyDType	DType which can represent any dtype.
enp.dtypes.AnyFloat	Generic float dtype (float32, float64, bfloat16,…).
enp.dtypes.AnyInt	Generic int dtype (int32, int64, uint8,…).
enp.dtypes.Casting	Casting mode when converting to array.
enp.dtypes.DType	DType wrapper.
enp.dtypes.NpDType	Raw numpy dtype.
enp.typing.Array
enp.typing.ArrayAliasMeta	Metaclass to create array aliases.
enp.typing.BoolArray
enp.typing.FloatArray
enp.typing.IntArray
enp.typing.PRNGKey
enp.typing.StrArray
enp.typing.bool_
enp.typing.complex128
enp.typing.complex64
enp.typing.f16
enp.typing.f32
enp.typing.f64
enp.typing.i16
enp.typing.i32
enp.typing.i64
enp.typing.i8
enp.typing.ui16
enp.typing.ui32
enp.typing.ui64
enp.typing.ui8


enp.angle_between	Compute angle between 2 vectors, unsigned.
enp.batch_dot	Dot product on the last dimension, with broadcasting support.
enp.check_and_normalize_arrays	Check and normalize arrays.
enp.compat.astype	`x.astype(dtype)`.
enp.compat.concat	`xnp.concatenate(x, axis=axis)`.
enp.compat.det	Like `np.linalg.det` but auto-support jnp, tnp, np.
enp.compat.dtype_np_to_torch	Returns the torch dtype for the given numpy dtype.
enp.compat.dtype_torch_to_np	Returns the numpy dtype for the given torch dtype.
enp.compat.expand_dims	`xnp.expand_dims(x, axis=axis)`.
enp.compat.inv	Like `np.linalg.inv` but auto-support jnp, tnp, np.
enp.compat.is_array_xnp	`isinstance(x, xnp.Array)`.
enp.compat.norm	Like `np.linalg.norm` but auto-support jnp, tnp, np.
enp.compat.round	`x.round()` for jnp, tnp, np, otrch.
enp.get_np_module	Returns the numpy module associated with the given array.
enp.interp	Linearly scale the given value by the given range.
enp.is_array	Returns `True` if array is np or `jnp` array.
enp.is_array_str	Returns True if the given array is a `str` array.
enp.is_dtype_str	Returns True if the dtype is `str`.
enp.linalg.normalize	Normalize the vector to the unit norm.
enp.normalize_bytes2str	Normalize `bytes` array to `str` (UTF-8).
enp.project_onto_plane	Project `u` onto the plane `n` (orthogonal vector).
enp.project_onto_vector	Project `u` onto `v`.


enp.compat.lazy
enp.lazy
enp.linalg.lazy
enp.tau	Mirror math.tau (PEP 628). See https://tauday.com/


enp.typing.ArrayLike	ArrayLike indicates that any `np.array` input is also supported. For example: `A…