autoray.autoray
#
AUTORAY - backend agnostic array operations.
Copyright 2019-2023 Johnnie Gray
Licensed under the Apache License, Version 2.0 (the “License”); you may not use this file except in compliance with the License. You may obtain a copy of the License at
Unless required by applicable law or agreed to in writing, software distributed under the License is distributed on an “AS IS” BASIS, WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. See the License for the specific language governing permissions and limitations under the License.
Module Contents#
Classes#
A singleton object to use as a placeholder in a pytree, for |
|
Compose an |
|
A class to mimic the syntax of using numpy directly. |
Functions#
|
Do function named |
|
This is the default backend dispatcher, used if no global backend has |
|
|
|
|
|
Return the universally set backend, if any. |
|
Set a default global backend. The argument |
|
Context manager for setting a default backend. The argument |
|
Register the name (and by default the module or submodule) of a custom |
|
Get the name of the library that defined the class of |
|
|
|
Infer which backend should be used for a function that takes multiple |
|
Choose a backend based on function name, arguments, and the |
|
|
|
Cached retrieval of correct function for backend, all the logic for |
|
Directly provide your own function. |
|
Register a new container type for use with |
The default function to determine if an object is a leaf. This simply |
|
|
An alternative leaf tester for addressing only arrays within trees. |
|
|
|
Map |
|
|
|
Iterate over all leaves in |
|
|
|
Apply |
|
Flatten |
|
Unflatten |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
Take a function consisting of multiple |
|
Get the shape of an array as a tuple of int. This should be preferred |
|
Get the number of dimensions of an array. This should be preferred to |
|
Get the size, or number of elements, of an array. This should be |
|
Array conjugate. |
|
Array transpose. |
|
Array Hermitian transpose. |
|
Array real part. |
|
Array imaginary part. |
|
Array reshaped. |
|
Turn string specifier |
Find string specifier |
|
|
Compute the minimal dtype sufficient for |
|
Cast array as type |
|
Get a numpy version of array |
|
|
|
|
Add ability to handle dtype keyword. |
|
|
Wrap a function to match the api of another according to a translation. |
|
|
|
|
|
Register a new dispatcher. |
|
Try to infer backend from first argument passed to function. |
|
Dispatcher for functions where first argument is a sequence. |
|
Dispatcher for handling einsum. |
|
There are cases when we want to take into account both backends of two |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
Attributes#
- autoray.autoray.do(fn, *args, like=None, **kwargs)[source]#
Do function named
fn
on(*args, **kwargs)
, peforming single dispatch to retrievefn
based on whichever library defines the class of theargs[0]
, or thelike
keyword argument if specified.Examples
Works on numpy arrays:
>>> import numpy as np >>> x_np = np.random.uniform(size=[5]) >>> y_np = do('sqrt', x_np) >>> y_np array([0.32464973, 0.90379787, 0.85037325, 0.88729814, 0.46768083])
>>> type(y_np) numpy.ndarray
Works on cupy arrays:
>>> import cupy as cp >>> x_cp = cp.random.uniform(size=[5]) >>> y_cp = do('sqrt', x_cp) >>> y_cp array([0.44541656, 0.88713113, 0.92626237, 0.64080557, 0.69620767])
>>> type(y_cp) cupy.core.core.ndarray
Works on tensorflow arrays:
>>> import tensorflow as tf >>> x_tf = tf.random.uniform(shape=[5]) >>> y_tf = do('sqrt', x_tf) >>> y_tf <tf.Tensor 'Sqrt_1:0' shape=(5,) dtype=float32>
>>> type(y_tf) tensorflow.python.framework.ops.Tensor
You get the idea.
For functions that don’t dispatch on the first argument you can use the
like
keyword:>>> do('eye', 3, like=x_tf) <tf.Tensor: id=91, shape=(3, 3), dtype=float32>
- autoray.autoray._default_infer_from_sig(fn, *args, **kwargs)[source]#
This is the default backend dispatcher, used if no global backend has been set. Hot swapping this function out as below avoids having to check manually for a global backend or worse, a thread aware global backend, on every call to
do
.
- autoray.autoray._global_backend#
- autoray.autoray._global_backends_threadaware#
- autoray.autoray._inferrers_threadaware#
- autoray.autoray._importing_thrid#
- autoray.autoray._backend_lock#
- autoray.autoray.get_backend(get_globally='auto')[source]#
Return the universally set backend, if any.
- Parameters:
get_globally ({"auto", False, True}, optional) –
Which backend to return:
True: return the globally set backend, if any.
False: return the backend set for the current thread, if any.
”auto”: return the globally set backend, if this thread is the thread that imported autoray. Otherwise return the backend set for the current thread, if any.
- Returns:
backend – The name of the backend, or None if no backend is set.
- Return type:
str or None
- autoray.autoray.set_backend(like, set_globally='auto')[source]#
Set a default global backend. The argument
like
can be an explicit backend name or anarray
.- Parameters:
like (str or array) – The backend to set. If an array, the backend of the array’s class will be set.
set_globally ({"auto", False, True}, optional) –
Whether to set the backend globally or for the current thread:
True: set the backend globally.
False: set the backend for the current thread.
”auto”: set the backend globally if this thread is the thread that imported autoray. Otherwise set the backend for the current thread.
Only one thread should ever call this function with
set_globally=True
, (by default this is importing thread).
- autoray.autoray.backend_like(like, set_globally='auto')[source]#
Context manager for setting a default backend. The argument
like
can be an explicit backend name or anarray
to infer it from.- Parameters:
like (str or array) – The backend to set. If an array, the backend of the array’s class will be set.
set_globally ({"auto", False, True}, optional) –
Whether to set the backend globally or for the current thread:
True: set the backend globally.
False: set the backend for the current thread.
”auto”: set the backend globally if this thread is the thread that imported autoray. Otherwise set the backend for the current thread.
Only one thread should ever call this function with
set_globally=True
, (by default this is importing thread).
- autoray.autoray._CUSTOM_BACKENDS#
- autoray.autoray.register_backend(cls, name)[source]#
Register the name (and by default the module or submodule) of a custom array class.
- Parameters:
cls (type) – The array class itself.
name (str) – The name of the backend that should be used for this class. By default this wil be assumed to be the location of the relevant functions for this class, but this can be overridden.
- autoray.autoray.infer_backend(array)[source]#
Get the name of the library that defined the class of
array
- unlessarray
is directly a subclass ofnumpy.ndarray
, in which case assumenumpy
is the desired backend.
- autoray.autoray.multi_class_priorities#
- autoray.autoray.infer_backend_multi(*arrays)[source]#
Infer which backend should be used for a function that takes multiple arguments. This assigns a priority to each backend, and returns the backend with the highest priority. By default, the priority is:
builtins
: -2numpy
: -1other backends: 0
autoray.lazy
: 1
I.e. when mixing with
numpy
, other array libraries are preferred, when mixing withautoray.lazy
,autoray.lazy
is preferred. This has quite low overhead due to caching.
- autoray.autoray.choose_backend(fn, *args, like=None, **kwargs)[source]#
Choose a backend based on function name, arguments, and the
like
keyword argument. The default, iflike
is not specified, is to infer the backend from the function call, the default of which is simply to use the first argument, if no custom dispatcher is found. Otherwise the backend is chosen based on thelike
argument - which can be an explicit backend name or an arbitrary object.
- autoray.autoray._BACKEND_ALIASES#
- autoray.autoray._MODULE_ALIASES#
- autoray.autoray._SUBMODULE_ALIASES#
- autoray.autoray._FUNC_ALIASES#
- autoray.autoray._CUSTOM_WRAPPERS#
- autoray.autoray._FUNCS#
- autoray.autoray._COMPOSED_FUNCTION_GENERATORS#
- autoray.autoray.get_lib_fn(backend, fn)[source]#
Cached retrieval of correct function for backend, all the logic for finding the correct funtion only runs the first time.
- Parameters:
backend (str) – The module defining the array class to dispatch on.
fn (str) – The function to retrieve.
- Return type:
callable
- autoray.autoray.register_function(backend, name, fn, wrap=False)[source]#
Directly provide your own function.
- Parameters:
backend (str) – The name of the backend to register the function for.
name (str) – Name of the function, e.g. ‘sum’ or ‘linalg.svd’.
fn (callable) – The function to register.
wrap (bool, optional) – Whether to wrap the old function like
fn(old_fn)
rather than directly supply the entire new function.
- autoray.autoray.TREE_MAP_REGISTRY#
- autoray.autoray.TREE_APPLY_REGISTRY#
- autoray.autoray.TREE_ITER_REGISTRY#
- autoray.autoray.tree_register_container(cls, mapper, iterator, applier)[source]#
Register a new container type for use with
tree_map
andtree_apply
.- Parameters:
cls (type) – The container type to register.
mapper (callable) – A function that takes
f
,tree
andis_leaf
and returns a new tree of typecls
withf
applied to all leaves.applier (callable) – A function that takes
f
,tree
andis_leaf
and appliesf
to all leaves intree
.
- autoray.autoray.IS_CONTAINER_CACHE#
- autoray.autoray.is_not_container(x)[source]#
The default function to determine if an object is a leaf. This simply checks if the object is an instance of any of the registered container types.
- autoray.autoray.is_array(x)[source]#
An alternative leaf tester for addressing only arrays within trees.
- autoray.autoray.TREE_MAPPER_CACHE#
- autoray.autoray.tree_map(f, tree, is_leaf=is_not_container)[source]#
Map
f
over all leaves intree
, returning a new pytree.- Parameters:
f (callable) – A function to apply to all leaves in
tree
.tree (pytree) – A nested sequence of tuples, lists, dicts and other objects.
is_leaf (callable) – A function to determine if an object is a leaf,
f
is only applied to objects for whichis_leaf(x)
returnsTrue
.
- Return type:
pytree
- autoray.autoray.TREE_ITER_CACHE#
- autoray.autoray.tree_iter(tree, is_leaf=is_not_container)[source]#
Iterate over all leaves in
tree
.- Parameters:
f (callable) – A function to apply to all leaves in
tree
.tree (pytree) – A nested sequence of tuples, lists, dicts and other objects.
is_leaf (callable) – A function to determine if an object is a leaf,
f
is only applied to objects for whichis_leaf(x)
returnsTrue
.
- autoray.autoray.TREE_APPLIER_CACHE#
- autoray.autoray.tree_apply(f, tree, is_leaf=is_not_container)[source]#
Apply
f
to all leaves intree
, no new pytree is built.- Parameters:
f (callable) – A function to apply to all leaves in
tree
.tree (pytree) – A nested sequence of tuples, lists, dicts and other objects.
is_leaf (callable) – A function to determine if an object is a leaf,
f
is only applied to objects for whichis_leaf(x)
returnsTrue
.
- class autoray.autoray.Leaf[source]#
A singleton object to use as a placeholder in a pytree, for unflattening.
- __slots__ = ()#
- autoray.autoray.LEAF#
- autoray.autoray.tree_flatten(tree, is_leaf=is_not_container, get_ref=False)[source]#
Flatten
tree
into a list of leaves.- Parameters:
tree (pytree) – A nested sequence of tuples, lists, dicts and other objects.
is_leaf (callable) – A function to determine if an object is a leaf, only objects for which
is_leaf(x)
returnsTrue
are returned in the flattened list.get_ref (bool) – If
True
, a reference tree is also returned which can be used to reconstruct the original tree from a flattened list.
- Returns:
objs (list) – The flattened list of leaf objects.
(ref_tree) (pytree) – If
get_ref
isTrue
, a reference tree, with leaves ofLeaf
, is returned which can be used to reconstruct the original tree.
- autoray.autoray.tree_unflatten(objs, tree, is_leaf=is_leaf_placeholder)[source]#
Unflatten
objs
into a pytree of the same structure astree
.- Parameters:
objs (sequence) – A sequence of objects to be unflattened into a pytree.
tree (pytree) – A nested sequence of tuples, lists, dicts and other objects, the objs will be inserted into a new pytree of the same structure.
is_leaf (callable) – A function to determine if an object is a leaf, only objects for which
is_leaf(x)
returnsTrue
will have the next item fromobjs
inserted. By default checks for theLeaf
object inserted bytree_flatten(..., get_ref=True)
.
- Return type:
pytree
- class autoray.autoray.Composed(fn, name=None)[source]#
Compose an
autoray.do
using function. See the main wrappercompose
.
- autoray.autoray.compose(fn, *, name=None)[source]#
Take a function consisting of multiple
autoray.do
calls and compose it into a new, single, named function, registered withautoray.do
.This creates a default implementation of this function for each new backend encountered without explicitly having to write each out, but also allows for specific implementations to be overridden for specific backends.
If the function takes a
backend
argument, it will be supplied with the backend name, to save having to re-choose the backend.Specific implementations can be provided by calling the
register
method of the composed function, or it can itself be used like a decorator:@compose def foo(x): ... @foo.register("numpy") @numba.njit def foo_numba(x): ...
- Parameters:
fn (callable) – The funtion to compose, and its default implementation.
name (str, optional) – The name of the composed function. If not provided, the name of the function will be used.
- autoray.autoray.shape(x)[source]#
Get the shape of an array as a tuple of int. This should be preferred to calling x.shape directly, as it:
Allows customization (e.g. for torch and aesara which return different types for shape - use @shape.register(backend) to customize the behavior from this default implementation).
Can be used on nested lists and tuples, without calling numpy.
- Parameters:
x (array_like) – The array to get the shape of. It can be an arbitrary nested list or tuple of arrays and scalars, but is assumed not to be ragged.
- Returns:
shape – The size of each dimension of the array.
- Return type:
tuple of int
- autoray.autoray.ndim(x)[source]#
Get the number of dimensions of an array. This should be preferred to calling x.ndim, since not all backends implement that, and it can also be called on nested lists and tuples.
- Parameters:
x (array_like) – The array to get the number of dimensions of. It can be an arbitrary nested list or tuple of arrays and scalars.
- Returns:
ndim
- Return type:
int
- autoray.autoray.size(x)[source]#
Get the size, or number of elements, of an array. This should be preferred to calling x.size, since not all backends implement that, and it can also be called on nested lists and tuples.
- Parameters:
x (array_like) – The array to get the size of. It can be an arbitrary nested list or tuple of arrays and scalars.
- Returns:
size
- Return type:
int
- autoray.autoray.to_backend_dtype(dtype_name, like)[source]#
Turn string specifier
dtype_name
into dtype of backendlike
.
- autoray.autoray._COMPLEX_DTYPES#
- autoray.autoray._DOUBLE_DTYPES#
- autoray.autoray._DTYPE_MAP#
- autoray.autoray.astype(x, dtype_name, **kwargs)[source]#
Cast array as type
dtype_name
- triesx.astype
first.
- autoray.autoray.with_dtype_wrapper(fn)[source]#
Add ability to handle dtype keyword. If not None, dtype should be specified as a string, otherwise conversion will happen regardless.
- autoray.autoray.translate_wrapper(fn, translator)[source]#
Wrap a function to match the api of another according to a translation. The
translator
entries in the form of an ordered dict should have entries like:(desired_kwarg: (backend_kwarg, default_value))
with the order defining the args of the function.
- autoray.autoray.register_dispatch(fun, dispatcher)[source]#
Register a new dispatcher.
This is useful in case the backend to be used by a function cannot be inferred from the first argument.
- autoray.autoray.default_dispatcher(*args, **kwargs)[source]#
Try to infer backend from first argument passed to function.
- autoray.autoray._DISPATCHERS#
- autoray.autoray.join_array_dispatcher(*args, **kwargs)[source]#
Dispatcher for functions where first argument is a sequence.
- autoray.autoray.einsum_dispatcher(*args, **_)[source]#
Dispatcher for handling einsum.
einsum can be called with a str equation as the first argument, or with ‘interleaved’ inputs. This dispatcher handles both cases and also takes into account all arrays.
- autoray.autoray.binary_dispatcher(*args, **_)[source]#
There are cases when we want to take into account both backends of two arguments, e.g. a lazy variable and a constant array.
- autoray.autoray._partial_functions#
- class autoray.autoray.NumpyMimic(submodule=None)[source]#
A class to mimic the syntax of using numpy directly.
- autoray.autoray.numpy#
- autoray.autoray.numpy_linalg#
- autoray.autoray.numpy_random#
- autoray.autoray._builtin_dtype_lookup#
- autoray.autoray._JAX_RANDOM_KEY#