Model Class Reference

CPP API: mio::isecir::Model Class Reference

CPP API

#include <model.h>

Collaboration diagram for mio::isecir::Model:

Public Member Functions
bool	check_constraints (ScalarType dt) const
	Checks constraints on model parameters and initial data. More...

ScalarType	get_global_support_max (ScalarType dt) const
	Getter for the global support_max, i.e. More...

int	get_initialization_method_compartments () const
	Returns the index of the automatically selected initialization method. More...

size_t	get_num_agegroups () const
	Getter for number of age groups. More...

int	get_state_flat_index (Eigen::Index state_idx, AgeGroup agegroup) const
	Returns a flat index for the TimeSeries populations which contains values for the InfectionStates. More...

int	get_transition_flat_index (Eigen::Index transition_idx, AgeGroup agegroup) const
	Returns a flat index for the TimeSeries transitions which contains values for the InfectionTransitions. More...

	Model (TimeSeries< ScalarType > &&transitions_init, CustomIndexArray< ScalarType, AgeGroup > N_init, CustomIndexArray< ScalarType, AgeGroup > deaths_init, size_t num_agegroups, CustomIndexArray< ScalarType, AgeGroup > total_confirmed_cases_init=CustomIndexArray< ScalarType, AgeGroup >())
	Constructor to create an IDE-SECIR model. More...

void	set_tol_for_support_max (ScalarType new_tol)
	Setter for the tolerance used to calculate the maximum support of the TransitionDistributions. More...

Public Attributes
ParameterSet	parameters {AgeGroup(m_num_agegroups)}
	ParameterSet of Model Parameters. More...

TimeSeries< ScalarType >	populations
	TimeSeries containing points of time and the corresponding number of. More...

CustomIndexArray< ScalarType, AgeGroup >	total_confirmed_cases
	CustomIndexArray that contains the total number of confirmed cases at time t0 for every AgeGroup. More...

TimeSeries< ScalarType >	transitions
	TimeSeries containing points of time and the corresponding number of individuals transitioning from. More...

Private Types
using	ParameterSet = Parameters

Private Member Functions
void	compute_compartment_from_flows (ScalarType dt, Eigen::Index idx_InfectionState, AgeGroup group, Eigen::Index idx_IncomingFlow, int idx_TransitionDistribution1, int idx_TransitionDistribution2=0)
	Compute the compartment specified in idx_InfectionState at the current time – only using historic flow values and disrespecting potential, previous compartment value. More...

void	compute_flow (Eigen::Index idx_InfectionTransitions, Eigen::Index idx_IncomingFlow, ScalarType dt, AgeGroup group)
	Computes size of a flow for the current last time value in transitions. More...

void	compute_flow (Eigen::Index idx_InfectionTransitions, Eigen::Index idx_IncomingFlow, ScalarType dt, Eigen::Index current_time_index, AgeGroup group)
	Computes size of a flow. More...

void	compute_forceofinfection (ScalarType dt, bool initialization=false)
	Computes force of infection for the current last time in transitions. More...

void	compute_susceptibles (ScalarType dt)
	Computes number of Susceptibles for the current last time in populations. More...

void	flows_current_timestep (ScalarType dt)
	Sets all required flows for the current last timestep in transitions. More...

void	initial_compute_compartments (ScalarType dt)
	Computes the values of compartments at initialization. More...

void	initial_compute_compartments_infection (ScalarType dt)
	Computes the values of the infection compartments subset at initialization. More...

void	set_transitiondistributions_derivative (ScalarType dt)
	Setter for the vector m_transitiondistributions_derivative that contains the approximated derivative for all TransitionDistributions for all necessary time points. More...

void	set_transitiondistributions_in_forceofinfection (ScalarType dt)
	Setter for the vector m_transitiondistributions_in_forceofinfection. More...

void	set_transitiondistributions_support_max (ScalarType dt)
	Setter for the vector m_transitiondistributions_support_max that contains the support_max for all TransitionDistributions. More...

void	update_compartment_from_flow (InfectionState infectionState, std::vector< InfectionTransition > const &IncomingFlows, std::vector< InfectionTransition > const &OutgoingFlows, AgeGroup group)
	Updates the values of one compartment using flows. More...

void	update_compartments ()
	Updates the values of all compartments except Susceptible at initialization. More...

Private Attributes
CustomIndexArray< ScalarType, AgeGroup >	m_forceofinfection
	Force of infection term needed for numerical scheme. More...

int	m_initialization_method {0}
	Gives the index of the method used for the initialization of the model. More...

CustomIndexArray< ScalarType, AgeGroup >	m_N
	Vector containing the total population size of the considered region for every AgeGroup. More...

size_t	m_num_agegroups
	Number of Age Groups. More...

ScalarType	m_tol {1e-10}
	Tolerance used to calculate the maximum support of the TransitionDistributions. More...

CustomIndexArray< std::vector< std::vector< ScalarType > >, AgeGroup >	m_transitiondistributions_derivative
	CustomIndexArray. More...

CustomIndexArray< std::vector< std::vector< ScalarType > >, AgeGroup >	m_transitiondistributions_in_forceofinfection
	CustomIndexArray. More...

CustomIndexArray< std::vector< ScalarType >, AgeGroup >	m_transitiondistributions_support_max
	CustomIndexArray. More...

Friends
template<typename EntryType >
IOResult< void >	set_initial_flows (Model &model, const ScalarType dt, const std::vector< EntryType > rki_data, const Date date, const CustomIndexArray< ScalarType, AgeGroup > scale_confirmed_cases)

class	Simulation

Member Typedef Documentation

◆ ParameterSet

using mio::isecir::Model::ParameterSet = Parameters

private

Constructor & Destructor Documentation

◆ Model()

mio::isecir::Model::Model	(	TimeSeries< ScalarType > &&	transitions_init,
		CustomIndexArray< ScalarType, AgeGroup >	N_init,
		CustomIndexArray< ScalarType, AgeGroup >	deaths_init,
		size_t	num_agegroups,
		CustomIndexArray< ScalarType, AgeGroup >	total_confirmed_cases_init = `CustomIndexArray<ScalarType, AgeGroup>()`
	)

Constructor to create an IDE-SECIR model.

Parameters

[in,out]	transitions_init	TimeSeries with the initial values of the number of individuals, which transit within one timestep dt from one compartment to another. Possible transitions are specified in InfectionTransitions. Considered time points should have the distance dt. The last time point determines the start time t0 of the simulation. The time history must reach a certain point in the past so that the simulation can be performed. A warning is displayed if the condition is violated.
[in]	N_init	A vector, containing the populations of the considered region, for every AgeGroup.
[in]	deaths_init	A vector, containing the total number of deaths at time t0, for every AgeGroup.
[in]	num_agegroups	The number of AgeGroups.
[in]	total_confirmed_cases_init	A vector, containing the total confirmed cases at time t0 can be set if it should be used for initialization, for every AgeGroup.

Member Function Documentation

◆ check_constraints()

bool mio::isecir::Model::check_constraints ( ScalarType dt ) const

Checks constraints on model parameters and initial data.

Returns: Returns true if one (or more) constraint(s) are not satisfied, otherwise false.

◆ compute_compartment_from_flows()

void mio::isecir::Model::compute_compartment_from_flows	(	ScalarType	dt,
		Eigen::Index	idx_InfectionState,
		AgeGroup	group,
		Eigen::Index	idx_IncomingFlow,
		int	idx_TransitionDistribution1,
		int	idx_TransitionDistribution2 = `0`
	)

private

Compute the compartment specified in idx_InfectionState at the current time – only using historic flow values and disrespecting potential, previous compartment value.

The computation is meaningful for all compartments except Susceptible, Recovered and #Death and mostly needed for initialization. For Susceptible, Recovered and Dead, use corresponding alternative functions.

Parameters

[in]	dt	Time discretization step size.
[in]	idx_InfectionState	Specifies the considered InfectionState
[in]	group	The AgeGroup for which we want to compute.
[in]	idx_IncomingFlow	Specifies the index of the incoming flow to InfectionState in transitions.
[in]	idx_TransitionDistribution1	Specifies the index of the first relevant TransitionDistribution, related to a flow from the considered InfectionState to any other InfectionState. This index is also used for related probability.
[in]	idx_TransitionDistribution2	Specifies the index of the second relevant TransitionDistribution, related to a flow from the considered InfectionState to any other InfectionState (in most cases to Recovered). Related probability is calculated via 1-probability[idx_TransitionDistribution1]. Sometimes the second index is not needed, e.g., if probability[idx_TransitionDistribution1]=1.

◆ compute_flow() [1/2]

void mio::isecir::Model::compute_flow	(	Eigen::Index	idx_InfectionTransitions,
		Eigen::Index	idx_IncomingFlow,
		ScalarType	dt,
		AgeGroup	group
	)

private

Computes size of a flow for the current last time value in transitions.

Computes size of one flow from InfectionTransition, specified in idx_InfectionTransitions, for the current last time value in transitions.

Parameters

[in]	idx_InfectionTransitions	Specifies the considered flow from InfectionTransition.
[in]	idx_IncomingFlow	Index of the flow in InfectionTransition, which goes to the considered starting compartment of the flow specified in idx_InfectionTransitions. Size of considered flow is calculated via the value of this incoming flow.
[in]	dt	Time step to compute flow for.
[in]	group	The Age Group for which we want to compute the flow.

◆ compute_flow() [2/2]

void mio::isecir::Model::compute_flow	(	Eigen::Index	idx_InfectionTransitions,
		Eigen::Index	idx_IncomingFlow,
		ScalarType	dt,
		Eigen::Index	current_time_index,
		AgeGroup	group
	)

private

Computes size of a flow.

Computes size of one flow from InfectionTransition, specified in idx_InfectionTransitions, for the time index current_time_index.

Parameters

[in]	idx_InfectionTransitions	Specifies the considered flow from InfectionTransition.
[in]	idx_IncomingFlow	Index of the flow in InfectionTransition, which goes to the considered starting compartment of the flow specified in idx_InfectionTransitions. Size of considered flow is calculated via the value of this incoming flow.
[in]	dt	Time step to compute flow for.
[in]	current_time_index	The time index the flow should be computed for.
[in]	group	The Age group for which we want to compute the flow.

◆ compute_forceofinfection()

void mio::isecir::Model::compute_forceofinfection	(	ScalarType	dt,
		bool	initialization = `false`
	)

private

Computes force of infection for the current last time in transitions.

Computed value is stored in m_forceofinfection.

Parameters

[in]	dt	Time discretization step size.
[in]	initialization	If true we are in the case of the initialization of the model. For this we need forceofinfection at time point t0-dt and not at the current last time (given by transitions) as in the other time steps.

◆ compute_susceptibles()

void mio::isecir::Model::compute_susceptibles ( ScalarType dt )

private

Computes number of Susceptibles for the current last time in populations.

Number is computed using previous number of Susceptibles and the force of infection (also from previous timestep). Number is stored at the matching index in populations.

Parameters

[in] dt Time discretization step size.

◆ flows_current_timestep()

void mio::isecir::Model::flows_current_timestep ( ScalarType dt )

private

Sets all required flows for the current last timestep in transitions.

New values are stored in transitions. Most values are computed via the function compute_flow().

Parameters

[in] dt Time step.

◆ get_global_support_max()

ScalarType mio::isecir::Model::get_global_support_max ( ScalarType dt ) const

Getter for the global support_max, i.e.

the maximum of support_max over all TransitionDistributions.

This determines how many initial values we need for the transitions. It may be possible to run the simulation with fewer time points than the value of the global support_max, but this number ensures that it is possible.

Parameters

[in] dt Time step size.

Returns: Global support_max.

◆ get_initialization_method_compartments()

int mio::isecir::Model::get_initialization_method_compartments ( ) const

inline

Returns the index of the automatically selected initialization method.

The initialization method is selected automatically based on the different values that need to be set beforehand. Infection compartments are always computed through historic flow. Initialization methods for Susceptible and Recovered are tested in the following order: 1.) If a positive number for the total number of confirmed cases is set, Recovered is set according to that value and Susceptibles are derived. 2.) If Susceptibles are set, Recovered will be derived. 3.) If Recovered are set directly, Susceptibles are derived. 4.) If none of the above is set with positive value, the force of infection is used as in Messina et al (2021) to set the Susceptibles.

Returns: Index representing the initialization method.

◆ get_num_agegroups()

size_t mio::isecir::Model::get_num_agegroups ( ) const

inline

Getter for number of age groups.

Returns: Returns number of age groups.

◆ get_state_flat_index()

int mio::isecir::Model::get_state_flat_index	(	Eigen::Index	state_idx,
		AgeGroup	agegroup
	)		const

inline

Returns a flat index for the TimeSeries populations which contains values for the InfectionStates.

In the TimeSeries we store a vector for each time point. In this vector we store values for the different InfectionStates for every AgeGroup. This function is used to get the right index in this vector for a specific AgeGroup and InfectionState.

Parameters

[in]	state_idx	Index at which InfectionState we want to evaluate.
[in]	agegroup	The agegroup for which we want to evaluate.

◆ get_transition_flat_index()

int mio::isecir::Model::get_transition_flat_index	(	Eigen::Index	transition_idx,
		AgeGroup	agegroup
	)		const

inline

Returns a flat index for the TimeSeries transitions which contains values for the InfectionTransitions.

In the TimeSeries we store a vector for each time point. In this vector we store values for the different InfectionTransitions for every AgeGroup. This function is used to get the right index in this vector for a specific AgeGroup and InfectionTransition.

Parameters

[in]	transition_idx	Index determining which InfectionTransition we want to evaluate.
[in]	agegroup	The agegroup for which we want to evaluate.

◆ initial_compute_compartments()

void mio::isecir::Model::initial_compute_compartments ( ScalarType dt )

private

Computes the values of compartments at initialization.

The initialization method is selected automatically based on the different values that need to be set beforehand. Infection compartments are always computed through historic flows. Initialization methods for Susceptible and Recovered are tested in the following order: 1.) If a positive number for the total number of confirmed cases is set, Recovered is set according to that value and Susceptibles are derived. 2.) If Susceptibles are set, Recovered will be derived. 3.) If Recovered are set directly, Susceptibles are derived. 4.) If none of the above is set with positive value, the force of infection is used as in Messina et al (2021) to set the Susceptibles.

The function calculate_compartments_initialization() is used in every method for the compartments Exposed, InfectedNoSymptoms, InfectedSymptoms, InfectedSevere and InfectedCritical. In addition, the force of infection is calculated for start time t_0.

Parameters

[in] dt Time discretization step size.

◆ initial_compute_compartments_infection()

void mio::isecir::Model::initial_compute_compartments_infection ( ScalarType dt )

private

Computes the values of the infection compartments subset at initialization.

The values for the compartments Exposed, InfectedNoSymptoms, InfectedSymptoms, InfectedSevere and InfectedCritical for time t_0 are calculated using the initial data in form of flows. Calculated values are stored in populations.

Parameters

[in] dt Time discretization step size.

◆ set_tol_for_support_max()

void mio::isecir::Model::set_tol_for_support_max ( ScalarType new_tol )

inline

Setter for the tolerance used to calculate the maximum support of the TransitionDistributions.

Parameters

[in] new_tol New tolerance.

◆ set_transitiondistributions_derivative()

void mio::isecir::Model::set_transitiondistributions_derivative ( ScalarType dt )

private

Setter for the vector m_transitiondistributions_derivative that contains the approximated derivative for all TransitionDistributions for all necessary time points.

The derivative is approximated using a backwards difference scheme. The number of necessary time points for each TransitionDistribution is determined using m_transitiondistributions_support_max.

Parameters

[in] dt Time step size.

◆ set_transitiondistributions_in_forceofinfection()

void mio::isecir::Model::set_transitiondistributions_in_forceofinfection ( ScalarType dt )

private

Setter for the vector m_transitiondistributions_in_forceofinfection.

When computing the force of infection, we evaluate the survival functions of the TransitionDistributions InfectedNoSymptomsToInfectedSymptoms, InfectedNoSymptomsToRecovered, InfectedSymptomsToInfectedSevere and InfectedSymptomsToRecovered, weighted by the corresponding TransitionProbabilities, at the same time points. Here, we compute these contributions to the force of infection term and store them in the vector m_transitiondistributions_in_forceofinfection so that we can access this vector for all following computations.

Parameters

[in] dt Time step size.

◆ set_transitiondistributions_support_max()

void mio::isecir::Model::set_transitiondistributions_support_max ( ScalarType dt )

private

Setter for the vector m_transitiondistributions_support_max that contains the support_max for all TransitionDistributions.

This determines how many summands are required when calculating flows, the force of infection or compartments.

Parameters

[in] dt Time step size.

◆ update_compartment_from_flow()

void mio::isecir::Model::update_compartment_from_flow	(	InfectionState	infectionState,
		std::vector< InfectionTransition > const &	IncomingFlows,
		std::vector< InfectionTransition > const &	OutgoingFlows,
		AgeGroup	group
	)

private

Updates the values of one compartment using flows.

New value is stored in populations. The value is calculated using the compartment size in the previous time step and the related flows of the current time step. Therefore the flows of the current time step should be calculated before using this function.

◆ update_compartments()

void mio::isecir::Model::update_compartments ( )

private

Updates the values of all compartments except Susceptible at initialization.

New values are stored in populations. The values are calculated using the compartment size in the previous time step and the related flows of the current time step. Therefore the flows of the current time step should be calculated before using this function.

Friends And Related Function Documentation

◆ set_initial_flows

template<typename EntryType >

IOResult<void> set_initial_flows	(	Model &	model,
		const ScalarType	dt,
		const std::vector< EntryType >	rki_data,
		const Date	date,
		const CustomIndexArray< ScalarType, AgeGroup >	scale_confirmed_cases
	)

friend

◆ Simulation

friend class Simulation

friend

Member Data Documentation

◆ m_forceofinfection

CustomIndexArray<ScalarType, AgeGroup> mio::isecir::Model::m_forceofinfection

private

Force of infection term needed for numerical scheme.

◆ m_initialization_method

int mio::isecir::Model::m_initialization_method {0}

private

Gives the index of the method used for the initialization of the model.

◆ m_N

CustomIndexArray<ScalarType, AgeGroup> mio::isecir::Model::m_N

private

Vector containing the total population size of the considered region for every AgeGroup.

◆ m_num_agegroups

size_t mio::isecir::Model::m_num_agegroups

private

Number of Age Groups.

◆ m_tol

ScalarType mio::isecir::Model::m_tol {1e-10}

private

Tolerance used to calculate the maximum support of the TransitionDistributions.

◆ m_transitiondistributions_derivative

CustomIndexArray<std::vector<std::vector<ScalarType> >, AgeGroup> mio::isecir::Model::m_transitiondistributions_derivative

private

CustomIndexArray.

◆ m_transitiondistributions_in_forceofinfection

CustomIndexArray<std::vector<std::vector<ScalarType> >, AgeGroup> mio::isecir::Model::m_transitiondistributions_in_forceofinfection

private

CustomIndexArray.

◆ m_transitiondistributions_support_max

CustomIndexArray<std::vector<ScalarType>, AgeGroup> mio::isecir::Model::m_transitiondistributions_support_max

private

CustomIndexArray.

◆ parameters

ParameterSet mio::isecir::Model::parameters {AgeGroup(m_num_agegroups)}

ParameterSet of Model Parameters.

◆ populations

TimeSeries<ScalarType> mio::isecir::Model::populations

TimeSeries containing points of time and the corresponding number of.

◆ total_confirmed_cases

CustomIndexArray<ScalarType, AgeGroup> mio::isecir::Model::total_confirmed_cases

CustomIndexArray that contains the total number of confirmed cases at time t0 for every AgeGroup.

◆ transitions

TimeSeries<ScalarType> mio::isecir::Model::transitions

TimeSeries containing points of time and the corresponding number of individuals transitioning from.

Model Class Reference

Public Member Functions

Public Attributes

Private Types

Private Member Functions

Private Attributes

Friends

Member Typedef Documentation

◆ ParameterSet

Constructor & Destructor Documentation

◆ Model()

Member Function Documentation

◆ check_constraints()

◆ compute_compartment_from_flows()

◆ compute_flow() [1/2]

◆ compute_flow() [2/2]

◆ compute_forceofinfection()

◆ compute_susceptibles()

◆ flows_current_timestep()

◆ get_global_support_max()

◆ get_initialization_method_compartments()

◆ get_num_agegroups()

◆ get_state_flat_index()

◆ get_transition_flat_index()

◆ initial_compute_compartments()

◆ initial_compute_compartments_infection()

◆ set_tol_for_support_max()

◆ set_transitiondistributions_derivative()

◆ set_transitiondistributions_in_forceofinfection()

◆ set_transitiondistributions_support_max()

◆ update_compartment_from_flow()

◆ update_compartments()

Friends And Related Function Documentation

◆ set_initial_flows

◆ Simulation

Member Data Documentation

◆ m_forceofinfection

◆ m_initialization_method

◆ m_N

◆ m_num_agegroups

◆ m_tol

◆ m_transitiondistributions_derivative

◆ m_transitiondistributions_in_forceofinfection

◆ m_transitiondistributions_support_max

◆ parameters

◆ populations

◆ total_confirmed_cases

◆ transitions