Skip to content

Package Overview

The main models implemented within the package are found in the GenerativeNetworkModel class.

Additionally, there are four main sub-packages. Below, we give the main functionality of each subpackage. See the full docs pages for the functions and classes available in each sub-package.

  1. gnm.generative_rules
  2. gnm.weight_criteria
  3. gnm.evaluation
  4. gnm.fitting
  5. gnm.defaults
  6. gnm.utils

model

There are two varieties of models implemented within the package:

  1. Binary models - these capture only the presence and absence of connections within a network, without capturing their strength
  2. Weighted models - these additionally capture the strengths of the connections within the network.

Both of these are implemented within the GenerativeNetworkModel class.

gnm.generative_rules

The gnm.generative_rules sub-package contains a collection of different generative rules that can be used to grow and develop the generative network.

gnm.weight_criteria

The gnm.weight_criteria sub-package contains a collection of optimisation criteria that can be used to update the weights of a weighted generative network model.

gnm.evaluation

The gnm.evaluation sub-package contains evaluation criteria which can be used assess the fit between a set of synthetic and real networks.

gnm.fitting

The gnm.fitting sub-package contains methods to fit parameters to a dataset and perform sweeps over parameters.

gnm.defaults

The gnm.defaults sub-package contains default values that can be used to run experiments out-of-the-box

gnm.utils

Finally, the gnm.utils sub-package contains other useful functionality for working with the toolbox, including functionality for computing statistics and various graph measures