TensorFunctionLibrary

Symbolica documentation for getting started, symbolic expressions, numerical evaluation, pattern matching, and APIs in Python and Rust.

`TensorFunctionLibrary`

class TensorFunctionLibrary

Numerical implementations of unary elementwise tensor functions.

Use with network execution to evaluate BroadcastFunction calls on numeric components. Complex conjugation is included by default. Symbolic function calls whose arguments are not numeric can remain symbolic.

Examples

from symbolica.community.tensor import TensorFunctionLibrary, BroadcastFunction
functions = TensorFunctionLibrary()
functions.register(BroadcastFunction("square"), lambda x: x * x)

Methods

Name	Description
`__new__`	Create a numerical function library with built-in conjugation.
`__repr__`	Return a summary of the registered elementwise tensor functions.
`register`	Register a numerical implementation for an elementwise function.

`new`

TensorFunctionLibrary.__new__() -> TensorFunctionLibrary

Create a numerical function library with built-in conjugation.

Returns

TensorFunctionLibrary Mutable library for BroadcastFunction implementations.

Examples

from symbolica.community.tensor import TensorFunctionLibrary
functions = TensorFunctionLibrary()

`repr`

TensorFunctionLibrary.__repr__() -> builtins.str

Return a summary of the registered elementwise tensor functions.

Examples

from symbolica.community import tensor as sp
functions = sp.TensorFunctionLibrary()
text = repr(functions)

`register`

TensorFunctionLibrary.register(
    function: BroadcastFunction,
    callback: typing.Callable[[float | complex], float | complex],
) -> None

Register a numerical implementation for an elementwise function.

Parameters

function (BroadcastFunction) Function to implement; an existing implementation is replaced.
callback (callable) Receives one float or complex value and returns a float or complex value. Results determine the numerical output type.

Returns

None Updates this function library.

Examples

from symbolica.community.tensor import Representation, TensorName, TensorExpression
space = Representation.euc(2)
A = TensorName("M")(space, space)
from symbolica.community.tensor import Tensor
tensor = Tensor.dense(A, [1.0, 2.0, 3.0, 4.0])
from symbolica.community.tensor import TensorFunctionLibrary, BroadcastFunction
square = BroadcastFunction("numeric_square")
functions = TensorFunctionLibrary()
functions.register(square, lambda value: value * value)
square(tensor).to_tensor(function_library=functions)[0, 1]
4.0