List of functions in StateSpacePartitions

Tree(; structured = false, arguments = (; minimum_probability = 0.01))

Description

A tree is a data structure that can be used to store partitions of a state space.

Keyword Arguments

• structured::Bool - Whether the tree is structured or not.
• arguments::NamedTuple - The arguments for the tree.

Returns

• Tree - Either a BinaryTree or an UnstructuredTree.

determine_partition(trajectory, tree_type::Tree{Val{false}, S}; override = false) where S

Decription

This function determines the partition of a trajectory into an unstructured tree. The tree structure is specified by the tree_type argument. The trajectory argument is a trajectory of states. The override keyword argument is a boolean indicating whether to override the truncation of the trajectory.

Arguments

• trajectory: a trajectory of states
• tree_type: a Tree type

Keyword Arguments

• override: a boolean indicating whether to override the truncation of the trajectory

Returns

• embedding: a Tree object

determine_partition(trajectory, tree_type::Tree{Val{true}, S}; override = false) where S

Decription

This function determines the partitioning of the state space into a binary tree. The tree is determined by recursively splitting the state space into two parts. The splitting is done by k-means clustering. The number of levels of the tree is determined by the levels keyword argument. If the trajectory is too long, it is truncated to roughly 1000 points per level. The override keyword argument can be used to override this behavior.

Arguments

• trajectory: a trajectory of states
• tree_type: a Tree type

Keyword Arguments

• override: a boolean indicating whether to override the truncation of the trajectory

Returns

• embedding: a Tree object