The state_containers submodule

This submodule contains functions and classes to store and convert mutation states used for training. It also contains a function to compute an independence model that can be used as a starting point for training a new MHN.

class mhn.training.state_containers.StateAgeContainer(int[:, :] mutation_data, double[:] ages)

Bases: StateContainer

This class is used as a wrapper like the StateContainer class, but also contains age information for each sample.

get_data_shape(self)

Returns:

Number of tumor samples and the number of genes stored in this object

Return type:

tuple

get_max_mutation_num(self)

Returns:

Maximum number of mutations present in a single sample.

class mhn.training.state_containers.StateContainer(int[:, :] mutation_data)

Bases: object

This class is used as a wrapper such that the C array containing the States can be referenced in a Python script.

It also makes sure that there aren’t more than 32 mutations present in a single sample as this would break the algorithms.

get_data_shape(self)

Returns:

Number of tumor samples and the number of genes stored in this object

Return type:

tuple

get_max_mutation_num(self)

Returns:

Maximum number of mutations present in a single sample.

mhn.training.state_containers.create_indep_model(StateContainer state_container)

Compute an independence model from the data stored in the StateContainer object, where the baseline hazard Theta_ii of each event is set to its empirical odds and the hazard ratios (off-diagonal entries) are set to exactly 1. The independence model is returned in logarithmic representation.

Parameters:

state_container (StateContainer) – Data used to compute the independence model.

Returns:

Independence model in logarithmic representation.

Return type:

np.ndarray