nnmt.lif.exp._sensitivity_measure

nnmt.lif.exp._sensitivity_measure(effective_connectivity, frequency, analysis_frequencies, resorted_eigenvalues_mask, eigenvalue_index)

Calculates sensitivity measure as in Eq. 7 in Bos et al. [2016].

Evaluates the sensitivity measure at a given frequency. By default, the effective connectivity is diagonalized and the eigenmode corresponding to the eigenvalue that is closest to the complex(1, 0) is chosen.

Another eigenmode can be specified by the parameter eigenvalue_index. The order of the eigenvalues can be specified by the resorted_eigenvalues_mask.

Parameters:

effective_connectivitynp.ndarray

Effective connectivity matrix.

frequencynp.float

Frequency at which the sensitivity is evaluated in Hz.

analysis_frequenciesnp.ndarray

Analysis frequencies.

resorted_eigenvalues_masknp.ndarray

Mapping from old to new indices (e.g. for resorting the eigenmodes) as obtained from _match_eigenvalues_across_frequencies. Shape : (num analysis freqs, num populations)

eigenvalue_indexint

Index specifying the eigenvalue and corresponding eigenmode for which the sensitivity measure is evaluated.

Returns:

dict

Dictionary containing the results of the sensitivity analysis.

critical_frequencynp.float

Frequency at which the sensitivity is evaluated in Hz.

critical_frequency_indexint

Index of critical_frequency in all analysis frequencies.

critical_eigenvaluenp.complex

Critical eigenvalue.

left_eigenvectornp.ndarray

Left eigenvector corresponding to the critical eigenvalue.

right_eigenvectornp.ndarray

Right eigenvector corresponding to the critical eigenvalue.

knp.complex

Vector point from critical eigenvalue to complex(1,0).

k_pernp.complex

Vector perpendiculat to k.

sensitivitynp.ndarray

Sensitivity measure. Shape : (num analysis freqs, num populations, num populations)

sensitivity_ampnp.ndarray

Projection of sensitivity measure that alters amplitude of peak in power spectrum. Shape : (num analysis freqs, num populations, num populations)

sensitivity_freqnp.ndarray

Projection of Sensitivity measure that alters frequency of peak power spectrum. Shape : (num analysis freqs, num populations, num populations)