symnum.codegen module#

Utility functions to generate NumPy function code.

class symnum.codegen.FunctionExpression(args, return_val)[source]#

Bases: object

Function defined by a symbolic expression and set of symbolic arguments.

Parameters:

  • args (tuple[Union[sympy.Expr, SymbolicArray], ...]) – Symbolic arguments to function.

  • return_val (Union[sympy.Expr, SymbolicArray]) – Symbolic expression in arguments corresponding to function return value.

__call__(*args)[source]#

Evaluate symbolic function.

Parameters:

*args (Union[sympy.Expr, SymbolicArray]) – Symbolic values to pass as positional arguments to function.

Returns:

Symbolic value corresponding to return value for function for passed symbolic arguments.

Return type:

Union[sympy.Expr, SymbolicArray]

args#
return_val#
class symnum.codegen.TupleArrayPrinter(settings=None)[source]#

Bases: NumPyPrinter

SymPy printer which uses nested tuples for array literals.

Rather than printing array-like objects as numpy.array calls with a nested list argument (which is not numba compatible) a nested tuple argument is used instead.

settings is passed to CodePrinter.__init__() module specifies the array module to use, currently ‘NumPy’, ‘CuPy’ or ‘JAX’.

doprint(expr, assign_to=None)#

Print the expression as code.

Parameters:

  • expr (Expression) – The expression to be printed.

  • assign_to (Symbol, string, MatrixSymbol, list of strings or Symbols (optional)) – If provided, the printed code will set the expression to a variable or multiple variables with the name or names given in assign_to.

emptyPrinter(expr)#
language = 'Python'#
modules = None#
property order#
parenthesize(item, level, strict=False)#
printmethod: str = '_pythoncode'#
reserved_words = {'False', 'None', 'True', 'and', 'as', 'assert', 'break', 'class', 'continue', 'def', 'del', 'elif', 'else', 'except', 'finally', 'for', 'from', 'global', 'if', 'import', 'in', 'is', 'lambda', 'nonlocal', 'not', 'or', 'pass', 'raise', 'return', 'try', 'while', 'with', 'yield'}#
classmethod set_global_settings(**settings)#

Set system-wide printing settings.

stringify(args, sep, level=0)#
tab = '    '#
symnum.codegen.generate_func(inputs, exprs, func_name, *, printer=None, numpy_module=None, namespace=None, jit=False, simplify=False)[source]#

Generate a Python function from symbolic expression(s).

Parameters:

  • inputs (Iterable[Union[sympy.Expr, SymbolicArray]]) – List of symbolic inputs to use as function arguments.

  • exprs (Iterable[Union[sympy.Expr, SymbolicArray, FunctionExpression]]) – List of symbolic expressions to use as function return values.

  • func_name (str) – Name to define generated function with.

  • printer (Optional[Printer]) – Instance of SymPy sympy.printing.printer.Printer (sub)class which produces string representations of symbolic expressions to be used in code for generated function. Defaults to an instance of TupleArrayPrinter, a subclass of sympy.printing.numpy.NumpyPrinter, which prints object argument to numpy.array() as a nested tuple rather than a nested list, which retains compatibility with numba which only supports tuple array literals.

  • numpy_module (Optional[ModuleType]) – Module implementing NumPy API to use in NumPy API calls in generated function. Defaults to numpy.

  • namespace (Optional[dict]) – Namespace to define generated function in. Default is to create a temporary module and define function in that namespace. Set namespace=globals() to define in current global namespace.

  • jit (bool) – If True enables just-in-time compilation of the generated function using Numba (requires numba package to be installed in the current Python environment). Default is jit=False.

  • simplify (bool) – Whether to try to simplify symbolic expressions in exprs before generating code.

symnum.codegen.numpify(*arg_shapes, **kwargs)[source]#

Decorator to convert SymPy symbolic array function to a NumPy function.

Parameters:

  • *arg_shapes (tuple[int]) – Variable length list of tuples and integers defining shapes of array arguments to func, e.g. if func takes two arguments x and y with x an array with shape (2, 2) and y an array with shape (2, 4, 3) the call signature would be of the form numpify((2, 2), (2, 4, 3))(func).

  • **kwargs – Any keyword arguments to generate_func().

Returns:

Decorator which takes a SymPy function of which accepts one or more SymbolicArray as arguments and returns a symbolic scalar or SymbolicArray value, and returns a corresponding NumPy function which accepts one or more numpy.ndarray arguments and returns a scalar or numpy.ndarray.

Return type:

Callable[[Callable[…, Union[sympy.Expr, SymbolicArray]]], Callable[…, Union[ScalarLike, NDArray]]]

symnum.codegen.numpify_func(sympy_func, *arg_shapes, **kwargs)[source]#

Generate a NumPy function from a SymPy symbolic array function.

Parameters:

  • sympy_func (Callable[..., Union[sympy.Expr, SymbolicArray]]) – Function which takes one or more SymbolicArray as arguments and returns a symbolic scalar expression or SymbolicArray value.

  • *arg_shapes (tuple[int]) – Variable length list of tuples or integers defining shapes of array arguments to func, e.g. if func takes two arguments x and y with x an array with shape (2, 2) and y an array with shape (2, 4, 3) the call signature would be of the form numpify_func(func, (2, 2), (2, 4, 3)).

  • **kwargs – Any keyword arguments to generate_func().

Returns:

Generated NumPy function which takes one or more numpy.ndarray arguments and return a scalar or numpy.ndarray value.

Return type:

Callable[…, Union[ScalarLike, NDArray]]

On this page
Show Source