rl_equation_solver.utilities.utilities

Collection of useful functions

Functions

build_adjacency_matrix(graph)	Build adjacency matrix from graph edges and labels
build_adjacency_matrix_custom(graph)	Build adjacency matrix from graph edges and labels
encode_onehot(labels)	Onehot encoding
get_json_graph(expr)	Make a graph plot of the internal representation of SymPy expression.
get_node_features(graph, feature_dict)	Get node features from feature dictionary.
get_node_labels(graph)	Get node labels from graph.
graph_walk(parent, expr, node_list, link_list)	Walk over the expression tree recursively creating nodes and links.
normalize(mx)	Row-normalize sparse matrix
pad_array(arr, length)	Pad array with zeros according the given length
parse_node_features(node_features, feature_dict)	Parse node features.
plot_state_as_graph(expr)	Make a graph plot of the internal representation of SymPy expression.
sparse_mx_to_torch_sparse_tensor(sparse_mx)	Convert a scipy sparse matrix to a torch sparse tensor.
to_graph(expr, feature_dict)	Make a graph plot of the internal representation of SymPy expression.
to_vec(expr, feature_dict[, state_dim])	Get state vector for given expression

Classes

Batch()	Graph Embedding or state vector Batch
Experience(state, action, next_state, reward)	Create new instance of Experience(state, action, next_state, reward)
GraphEmbedding(graph, n_observations, ...)	Graph embedding class for embedding node features in matrix of fixed sizes
Id()	A helper class for autoincrementing node numbers.
Node(label, expr_id)	Represents a single operation or atomic argument.
ReplayMemory(capacity)	Stores the Experience Replay buffer
VectorEmbedding(vector, n_observations, device)	Vector embedding class for embedding feature vector in vector of fixed size