nnmt.models.Microcircuit

class nnmt.models.Microcircuit(network_params=None, analysis_params=None, file=None)

The Potjans and Diesmann microcircuit model.

See Potjans and Diesmann [2012] for details regarding the model. In short, it is a four-layer (2/3, 4, 5, 6) network model with a population of excitatory (E) and inhibitory (I) neurons of leaky integrate-and-fire neurons with exponential synapses in each layer. The inhibitory synaptic weights are g times as strong as the excitatory synaptic weights. The weights between all populations are equally strong, except for layer 4E to layer 2/3E, where the excitatory weights are twice as strong.

Given the parameter yaml files, the network model calculates the dependend parameters. It converts the weights from pA to mV, calculates the weight matrix, calculates relative thresholds, and the analysis frequencies.

The NNMT repository contains an example providing the yaml parameter files with all the parameters that need to be defined to use this model.

Parameters:

network_params[str | dict]

Network parameters yaml file name or dictionary including:

• Cfloat

  Membrane capacitance in pF.

• K_extnp.array

  Number of external in-degrees.

• V_th_abs[float | np.array]

  Absolute threshold potential in mV.

• V_0_abs[float | np.array]

  Absolute reset potential in mV.

• d_efloat

  Mean delay of excitatory connections in ms.

• d_e_sdfloat

  Standard deviation of delay of excitatory connections in ms.

• d_ifloat

  Mean delay of inhibitory connections in ms.

• d_i_sd

  Standard deviation of delay of inhibitory connections in ms.

• gfloat

  Ratio of inhibitory to excitatory synaptic weights.

• populationslist of strings

  Names of different populations.

• tau_sfloat

  Synaptic time constant in ms.

• wfloat

  Amplitude of excitatory post synaptic current in pA.

• w_ext: float

  Amplitude of external excitatory post synaptic current in pA.

analysis_params[str | dict]

Analysis parameters yaml file name or dictionary including:

• dffloat

  Step size between two analysis frequencies.

• f_minfloat

  Minimal analysis frequency.

• f_maxfloat

  Maximal analysis frequency.

• dkfloat

  Step size between two analysis wavenumber.

• k_minfloat

  Minimum analysis wavenumber.

• k_max

  Maximum analysis wavenumber.

See also

nnmt.models.Network

Parent class defining all arguments, attributes, and methods.