DigitalContactTracing

A class that implements digital contact tracing on a real contact network.

The class loads an existing network and simulates the spread of a virus on the network, based on the characteristics of an infectious disease. At the same time, a digital contact tracing policy is implemented to try to contain the spread of the virus by enforcing isolation and quarantine, depending on the policy specifications. The class keeps track of a number of relevant quantities (mainly, tracing efficacy and histories of quarantined individuals).

Attributes of the class are listed below:

ATTRIBUTES	Type	Description
I	dict	details of infected people
infected	list	infected people
isolated	list	isolated people
quarantined	list	quarantined people
symptomatic	list	symptomatic people
eT	list	tracing effectivity, per time instant
temporal_gap	float	temporal gap between static networks
memory_contacts	int	tracing memory
max_time_quar	float	quarantine duration
contacts	list	contacts of each node
sym_t	list	symptomatic people, full history
iso_t	list	isolated people, full history
act_inf_t	list	infected people, full history
q_t	list	number of quarantined, full history
q_t_i	list	number of wrongly quarantined, full history
Q_list	list	quarantined, full history
Qi_lis	list	wrongly quarantined, full history
sympt	float	wrongly fraction of symptomatic individuals
test	float	wrongly fraction of asymptomatics who are detected via random testing
eps_I	float	wrongly isolation effectivity
filter_rssi	float	wrongly RSSI threshold of the digital tracing policy
filter_duration	float	wrongly duration threshold of the digital tracing policy
graphs	list	snapshots of the temporal graph
beta_t	float	parameter defining the infectiousness probability
use_rssi	bool	flag to decide if the simulation is on a SocioPattern dataset
Y_i_nodes	list	initially infected nodes
NC_nodes	list	nodes who do not use the app

Methods of the class are listed below:

METHODS	Description
spread_infection(graph, node, new_infected, current_time)	Propagates the infection from an infected node to its neighbors.
enforce_policy(current_time,node,in_quarantine)	Updates the state of symptomatic people.
inizialize_contacts(graph)	Initialize the contacts from the temporal graph.
inizialize_infected_time0()	Initialize the status of the initial infected people.
policy(graph, node)	Implements a policy on a node in a graph.
simulate()	Runs the simulation.
update_contacts(graph)	Updates the list of traced contacts.
update_quarantined(current_time)	Update the list of quarantined people.
update_infected( current_time, graph, new_infected)	Updates the state of the infected nodes.
check_quarantined( current_time)	Check the status of quarantined nodes.

spread_infection

spread_infection(graph, node, new_infected, current_time)

Propagates the infection from an infected node to its neighbors.

The method loops over the neighbors of an infected node and selectively propagates the infection (i.e., add the neighbors to the list of infected nodes). To decide if the infection is propagated or not, the method checks the duration and proximity of a contact and the infection probability beta_data.

INPUT

• graph - networkx.classes.graph.Graph snapshots of the temporal graph
• node - the id of the node under consideration
• new_infected - list of pearson newely infected
• current_time - float the absolute time since the beginning of the simulation

enforce_policy

enforce_policy(current_time,node,in_quarantine)

Update the state of a symptomatic node and quarantine its contacts.

The method updates the state of a node which is found infected, i.e., it is isolated and its contacts are quarantined. First, the node is added to the list of isolated nodes and removed from the list of infected (if it is not quarantined) or from the list of quarantined (if it is quarantined). Second, if the node is adopting the app, the list of its past contacts which were 'at risk' is processed and each node is quarantined. Third, the efficacy of this tracing step is computed and appended to the global list self.eTt.

INPUT

• node - the id of the node under consideration
• in_quarantine - boolean flag that specify if the pearson is already qurantained and he/she shows simptoms
• current_time - float the absolute time since the beginning of the simulation

inizialize_contacts

inizialize_contacts(graph)

Initialize the contacts from the temporal graph.

The method creates a list where the element at position idx is the list of contacts of the node idx. Each of these lists is initally empty.

INPUT

• graph - networkx.classes.graph.Graph snapshots of the temporal graph

inizialize_infected_time0

inizialize_infected_time0(graph)

Initialize the status of the initial infected people.

The method adds the people contained in self.Y_i_nodes to the list I, and defines their infectiousness properties.

policy

policy(graph, node)

Implements a policy on a node in a graph.

The method gets the neighbors of the node, and for each neighbor it applies the policy to decide if it is 'at risk' or not.

INPUT

• graph - networkx.classes.graph.Graph snapshots of the temporal graph
• node - the id of the node under consideration

simulate

simulate()

Run the simulation.

The method runs the simulation on the temporal network.

OUTPUT

• eT: list tracing effectivity, per time instant
• sym_t: list symptomatic people, full history
• iso_t: list isolated people, full history
• act_inf_t: list infected people, full history
• q_t: list number of quarantined, full history
• q_t_i: list number of wrongly quarantined, full history
• Q_nb: list number of distict elements in self.Q_list
• Qi_nb: list number of distinct elements in self.Qi_list
• I: dict details of infected people

update_contacts

update_contacts(current_time)

Update the list of quarantined people

INPUT

• current_time - the absolute time since the beginning of the simulation

update_infected

update_infected( current_time, graph, new_infected)

Updates the state of the infected nodes.

The method updates the state of each infected node by advancing in time its information and by checking if it has become symptomatic. Moreover, an infected node may be isolated according to the isolation efficiency: If it is not isolated, it spread the infection to its neighbors; If it is isolated, the tracing policy is enforced on its contacts.

INPUT

• current_time - float the absolute time since the beginning of the simulation
• graph - networkx.classes.graph.Graph snapshots of the temporal graph
• new_infected - list of pearson newely infected

check_quarantined

check_quarantined(current_time)

Check the status of quarantined nodes.

The method checks if a nodes becomes symptomatic while in quarantine, and in this case the tracing policy is enforced on its contacts.

INPUT

• current_time - float the absolute time since the beginning of the simulation

UTILITIES

store_real_time

store_real_time(res,PARAMETERS,filter_rssi,filter_duration,eps_I)

Store the results of the simulation at the end of each single iteration. If the file does not exists, it create the file, otherwise it append the current results to the file.

INPUT

• res - list that contains the output of the simulation
• PARAMETERS - dict containing all the parameters of the simulation (se the examples Digital-Contact-Tracing on SocioPattern.ipynb and Digital-Contact-Tracing on DTU.ipynb)
• filter_rssi - float wrongly RSSI threshold of the digital tracing policy
• filter_duration - float wrongly duration threshold of the digital tracing policy
• eps_I - float wrongly isolation effectivity

load_results

load_results(path,file,eps_I,filter_rssi,filter_duration)

Load stored results

INPUT

• path - string representing the path in which results are stored
• file - string specifing the file to load. One of the follow:
  • q_t number of quarantined
  • q_t_i number of wrongly quarantined
  • Q_list list of quarantined
  • Qi_lis list of wrongly quarantined
  • I dict with details of infected people
  • sym_t list of symptomatic people
  • iso_t list of isolated people
  • act_inf_t list of infected people
• eps_I - float wrongly isolation effectivity
• filter_rssi - float wrongly RSSI threshold of the digital tracing policy
• filter_duration - float wrongly duration threshold of the digital tracing policy

OUTPUT

• An array containing the loaded results.