connectome_manipulator.model_building.conn_prob_adj

Module for building (deterministic) connection probability models based on adjacency matrices

connectome_manipulator.model_building.conn_prob_adj.build(adj_mat, src_node_ids, tgt_node_ids, inverted=False, **_)[source]

Builds a (deterministic) connection probability model of type ConnProbAdjModel from an adjacency matrix (i.e., returning probabilities 0.0 or 1.0 only).

Parameters:

  • adj_mat (scipy.sparse.csc_matrix) – Sparse adjacency matrix with boolean entries (i.e., True…connection, False…no connection)

  • src_node_ids (list-like) – List of source (pre-synaptic) neuron IDs

  • tgt_node_ids (list-like) – List of target (post-synaptic) neuron IDs

  • inverted (bool) – Flag for interpreting the boolean matrix entries in an inverted way (i.e., True…no connection, False…connection)

Returns:

Resulting adjacency model

Return type:

connectome_manipulator.model_building.model_types.ConnProbAdjModel

connectome_manipulator.model_building.conn_prob_adj.extract(circuit, sel_src=None, sel_dest=None, edges_popul_name=None, CV_dict=None, **_)[source]

Extracts adjacency matrix between selected src/dest neurons.

Parameters:

  • circuit (bluepysnap.Circuit) – Input circuit

  • sel_src (str/list-like/dict) – Source (pre-synaptic) neuron selection

  • sel_dest (str/list-like/dict) – Target (post-synaptic) neuron selection

  • edges_popul_name (str) – Name of SONATA egdes population to extract data from

  • CV_dict (dict) – Cross-validation dictionary - Not supported

Returns:

Dictionary containing the extracted adjacency matrix and source/target node ids

Return type:

dict

connectome_manipulator.model_building.conn_prob_adj.plot(out_dir, model, **_)[source]

Visualizes the adjacency model.

Parameters:

  • out_dir (str) – Path to output directory where the results figures will be stored

  • model (connectome_manipulator.model_building.model_types.ConnProbAdjModel) – Adjacency model, as returned by build()