calculate_contact_matrix#

calculate_contact_matrix(population, density_or_frequency='density', layer='H')[source]#

Calculate the symmetric age-specific contact matrix from the connections for all people in the population. density_or_frequency sets the type of contact matrix calculated.

When density_or_frequency is set to ‘frequency’ each person is assumed to have a fixed amount of contact with others they are connected to in a setting so each person will split their contact amount equally among their connections. This means that if a person has links to 4 other individuals then 1/4 will be added to the matrix element matrix[age_i][age_j] for each link, where age_i is the age of the individual and age_j is the age of the contact. This follows the mass action principle such that increased density or number of people a person is in contact with leads to decreased per-link or connection contact rate.

When density_or_frequency is set to ‘density’ the amount of contact each person has with others scales with the number of people they are connected to. This means that a person with connections to 4 individuals has a higher total contact rate than a person with connection to 3 individuals. For this definition if a person has links to 4 other individuals then 1 will be added to the matrix element matrix[age_i][age_j] for each contact. This follows the ‘pseudo’mass action principle such that the per-link or connection contact rate is constant.

Parameters:

  • population (dict) – A dictionary of a population with attributes.

  • density_or_frequency (str) – option for the type of contact matrix calculated.

  • layer (str) – name of the physial contact setting, see notes.

Returns:

Symmetric age specific contact matrix.

Return type:

np.ndarray

Note

H for households, S for schools, W for workplaces, C for community or other, and ‘LTCF’ for long term care facilities.