geneticengine.representations.tree.initializations

Module Contents

Functions

apply_metahandler(r, g, receiver, new_symbol, depth, ...)	This method applies a metahandler to use a custom generator for things
grow_method(r, g, depth[, starting_symbol])	Implements the standard Grow tree-initialization method, where trees are
full_method(r, g, depth[, starting_symbol])	Full tree-initialization method.
pi_grow_method(r, g, depth[, starting_symbol])	PI Grow tree-initialization method.
random_list(r, receiver, new_symbol, depth, ty, ctx[, ...])
expand_node(r, g, new_symbol, filter_choices, ...)	Creates a random node of a given type (starting_symbol).
mk_save_init(starting_symbol, receiver)	Saves a child as a member of the parent node.

Attributes

InitializationMethodType
T

geneticengine.representations.tree.initializations.apply_metahandler(r, g, receiver, new_symbol, depth, ty, context)

This method applies a metahandler to use a custom generator for things of a given type.

As an example, AnnotatedType[int, IntRange(3,10)] will use the IntRange.generate(r, recursive_generator). The generator is the annotation on the type (“__metadata__”).

Parameters:

• r (geneticengine.random.sources.RandomSource)

• g (geneticengine.grammar.grammar.Grammar)

• depth (int)

• ty (type[Any])

• context (dict[str, str])

Return type:

Any

geneticengine.representations.tree.initializations.InitializationMethodType

geneticengine.representations.tree.initializations.grow_method(r, g, depth, starting_symbol=int)

Implements the standard Grow tree-initialization method, where trees are naturally grown from the grammar.

Parameters:

• r (geneticengine.random.sources.RandomSource)

• g (geneticengine.grammar.grammar.Grammar)

• depth (int)

• starting_symbol (type[Any])

geneticengine.representations.tree.initializations.full_method(r, g, depth, starting_symbol=int)

Full tree-initialization method.

Trees are grown from the grammar with all branches as deep as possible, making full trees.

Parameters:

• r (geneticengine.random.sources.RandomSource)

• g (geneticengine.grammar.grammar.Grammar)

• depth (int)

• starting_symbol (type[Any])

geneticengine.representations.tree.initializations.pi_grow_method(r, g, depth, starting_symbol=int)

PI Grow tree-initialization method.

(http://ncra.ucd.ie/papers/Exploring%20Position%20Independent%20Initialisation%20in%20Grammatical.%20Evolution.pdf), where trees are grown to have at least one branchas deep as possible.

Parameters:

• r (geneticengine.random.sources.RandomSource)

• g (geneticengine.grammar.grammar.Grammar)

• depth (int)

• starting_symbol (type[Any])

geneticengine.representations.tree.initializations.T

geneticengine.representations.tree.initializations.random_list(r, receiver, new_symbol, depth, ty, ctx, prod='')

Parameters:

• r (geneticengine.random.sources.RandomSource)

• depth (int)

• ty (type[list[T]])

• ctx (dict[str, str])

• prod (str)

geneticengine.representations.tree.initializations.expand_node(r, g, new_symbol, filter_choices, receiver, depth, starting_symbol, id, ctx)

Creates a random node of a given type (starting_symbol).

Parameters:

• r (geneticengine.random.sources.RandomSource)

• g (geneticengine.grammar.grammar.Grammar)

• starting_symbol (type)

• id (str)

• ctx (dict[str, str])

Return type:

Any

geneticengine.representations.tree.initializations.mk_save_init(starting_symbol, receiver)

Saves a child as a member of the parent node.

Parameters:

• starting_symbol (Any)

• receiver (Callable)