Compare distributions of network properties

braindynamics_starprotocol.network.compare.compare_ews_Cliffs_delta(A: ndarray, B: ndarray) → tuple[source]

Function that computes Cliff’s delta metric for two sets of edge weights (EWD samples). Because the EWD is a bimodal distribution, we calculate separately for the negative and for the positive parts of the distribution.

Parameters:

  • A (np.ndarray) – First array of edge weights with shape: <number of edges>, <number of trials>

  • B (np.ndarray) – Second array of edge weights with shape: <number of edges>, <number of trials>

Raises:

ValueError – Length of the first dimension (number of edges) don’t match in the two arrays.

Returns:

Two arrays representing Cliff’s delta metric of the positive and negative parts, for each edge.

Return type:

tuple

braindynamics_starprotocol.network.compare.compare_nds_Cliffs_delta(A: ndarray, B: ndarray) → ndarray[source]

Function that computes Cliff’s delta metric for two sets of node distances (NDD values).

Parameters:

  • A (np.ndarray) – First array of node distances with shape: <number of nodes>, (<number of nodes> - 1)*<total number of trials 1>.

  • B (np.ndarray) – Second array of node distances with shape: <number of nodes>, (<number of nodes> - 1)*<total number of trials 2>.

Raises:

ValueError – Length of first dimension (number of nodes) don’t match in the two arrays.

Returns:

Array representing Cliff’s delta metric for each node.

Return type:

np.ndarray

braindynamics_starprotocol.network.compare.compare_news_Cliffs_delta(A: ndarray, B: ndarray) → tuple[source]

Function that computes Cliff’s delta metric for two sets of node edge weights (N-EWD values). Because the N-EWD is a bimodal distribution, we calculate separately for the negative and for the positive parts of the distribution.

Parameters:

  • A (np.ndarray) – First array of node distances with shape: <number of nodes>, (<number of nodes> - 1)*<total number of trials 1>.

  • B (np.ndarray) – Second array of node distances with shape: <number of nodes>, (<number of nodes> - 1)*<total number of trials 2>.

Raises:

ValueError – Length of first dimension (number of nodes) don’t match in the two arrays.

Returns:

Two arrays representing Cliff’s delta metric of the positive and negative parts, for each node.

Return type:

tuple