library(diseasy)
#> Loading required package: diseasystore
#>
#> Attaching package: 'diseasy'
#> The following objects are masked from 'package:diseasystore':
#>
#> %.%, diseasyoptionIntroduction
The DiseasyActivity is the module responsible for the
implementation of contact matrices and societal restrictions.
Configuring the module
The module can be initialized without setting any parameters.
act <- DiseasyActivity$new()By default, the module initializes with no activity scenario, and produces empty outputs. (See the interfacing section for ways to get metrics from the module)
act$get_scenario_openness()
#> list()To use the module, we must define the activity scenario. In the example below, we configure the module to use Danish activity units and the “contact basis” for Denmark (i.e basic contact matrices and population information).
NOTE: the given contact basis uses counts of contacts between
individuals rather than rates. Having contacts as counts is important if
you want to project the contact matrices into different age groups. (see
?contact_basis for more details on how the
contact_basis is constructed)
act$set_activity_units(dk_activity_units)
act$set_contact_basis(contact_basis$DK)
# This configuration can also be done as part of the constructor
# act <- DiseasyActivity$new(activity_units = dk_activity_units
# contact_basis = contact_basis$DK)Defining an activity scenario
Once activity_units and contact_basis is
supplied, we can configure the activity scenario. In this example we
have a single activity defined (the baseline activity)
starting from 2023-01-01.
act$change_activity(date = as.Date("2020-01-01"), opening = "baseline")To get an overview of the loaded activities, we can use the
$get_scenario_activities() method.
act$get_scenario_activities()
#> $`2020-01-01`
#> [1] "baseline"Calls to $change_activity are cumulative, which means
that we can expand our existing scenario:
act$change_activity(date = as.Date("2020-03-11"), opening = "lockdown_2020", closing = "baseline")
# NOTE: we also have to "close" the "baseline" activity unit. The activity described by this unit is
# no longer in effect while the "lockdown_2020" activity unit is active.
act$get_scenario_activities()
#> $`2020-01-01`
#> [1] "baseline"
#>
#> $`2020-03-11`
#> [1] "lockdown_2020"If we try to add the same activity_unit while it is
already open, the module will give an error:
act$change_activity(date = as.Date("2020-03-11"), opening = "lockdown_2020")
#> Error in act$change_activity(date = as.Date("2020-03-11"), opening = "lockdown_2020"):
#> Some of lockdown_2020 are already open!By default, activity units are added with a risk modifier of 1. The module includes the option to set an overall risk associated with each of the 4 types of activity (“home”, “work”, “school”, “other”).
Having the risk as a separate modifier, allows the activity units to describe the level of contacts between individuals while the risk modifier can describe risk-mitigation measures that effect these contacts.
To modify the associated risk, we use the $change_risk()
method.
# Mitigation measures in workplaces reduce the risk
act$change_risk(date = as.Date("2020-03-11"), type = "work", risk = 0.45)
act$risk_matrix
#> 2020-01-01 2020-03-11
#> home 1 1.00
#> work 1 0.45
#> school 1 1.00
#> other 1 1.00To create a new activity scenario with an active instance of the
module, we can use the $reset_scenario() method:
act$reset_scenario()
act$get_scenario_activities()
#> list()The scenario can also be loaded directly from a
data.frame such as the
dk_reference_scenario
print(dk_reference_scenario)
#> # A tibble: 301 × 4
#> date opening closing social_distance_work
#> <date> <chr> <chr> <dbl>
#> 1 2020-01-01 baseline NA 1
#> 2 2020-03-11 NA baseline 0.45
#> 3 2020-03-11 lockdown_2020 NA 0.45
#> 4 2020-04-15 secondary_education_phase_1_2020 NA 0.45
#> 5 2020-04-15 private_companies_phase_1_2020 NA 0.45
#> # ℹ 296 more rows
act$change_activity(dk_reference_scenario)
activities <- act$get_scenario_activities()
# NOTE: The full list of activities is too long to print in this vignette
activities[1:3]
#> $`2020-01-01`
#> [1] "baseline"
#>
#> $`2020-03-11`
#> [1] "lockdown_2020"
#>
#> $`2020-04-15`
#> [1] "lockdown_2020" "secondary_education_phase_1_2020"
#> [3] "private_companies_phase_1_2020" "daycares_and_schools_phase_1_2020"The loaded scenario can be “cropped” through the
$crop_scenario() method.
act$crop_scenario(as.Date("2020-03-01"), as.Date("2020-04-15"))
act$get_scenario_activities()
#> $`2020-03-01`
#> [1] "baseline"
#>
#> $`2020-03-11`
#> [1] "lockdown_2020"
#>
#> $`2020-04-15`
#> [1] "lockdown_2020" "secondary_education_phase_1_2020"
#> [3] "private_companies_phase_1_2020" "daycares_and_schools_phase_1_2020"
# NOTE: The full list of activities in the scenario is shown here
# NOTE: The start date of the "baseline" activity is now updated to the given dateInterfacing with the activity scenario
For these examples, we reduce the activity scenario to just a single activity:
act$reset_scenario()
act$change_activity(date = as.Date("2020-03-11"), opening = "baseline")With the activity scenario loaded, we query the module in different ways to get contact metrics from the module for the scenario.
Contact matrices
Several disease models uses contact matrices to structure the populations.
To retrieve the contact matrices from the module, we use the
$get_scenario_contacts() method:
# We can project the matrices into custom age-groups by providing the "age_cuts_lower" argument
act$get_scenario_contacts(age_cuts_lower = c(0, 60))
#> $`2020-03-11`
#> $`2020-03-11`$home
#> 00-59 60+
#> 00-59 0.0025 0.00050
#> 60+ 0.0005 0.00025
#>
#> $`2020-03-11`$work
#> 00-59 60+
#> 00-59 1.5e-03 7.5e-05
#> 60+ 7.5e-05 5.6e-06
#>
#> $`2020-03-11`$school
#> 00-59 60+
#> 00-59 1.2e-03 3.1e-06
#> 60+ 3.1e-06 2.6e-08
#>
#> $`2020-03-11`$other
#> 00-59 60+
#> 00-59 0.00246 0.00037
#> 60+ 0.00037 0.00027The DiseasyActivity module provides separate matrices
for the four arenas: “home”, “school”, “work”, and “other”. These
matrices can be combined by supplying a weights argument to
the $get_scenario_contacts() method.
act$get_scenario_contacts(age_cuts_lower = c(0, 20, 40, 60, 80), weights = c(1, 1, 1, 1))
#> $`2020-03-11`
#> 00-19 20-39 40-59 60-79 80+
#> 00-19 0.00174 5.4e-04 4.7e-04 0.00018 3.0e-05
#> 20-39 0.00054 1.3e-03 7.9e-04 0.00029 3.6e-05
#> 40-59 0.00047 7.9e-04 1.1e-03 0.00036 5.7e-05
#> 60-79 0.00018 2.9e-04 3.6e-04 0.00042 5.0e-05
#> 80+ 0.00003 3.6e-05 5.7e-05 0.00005 9.6e-06Societal openness
A simpler metric for societal activity is the degree of “openness”, i.e. degree to which their activity is limited relative to an open society.
This value ranges from 1 (normal activity) to 0 (no activity).
act$get_scenario_openness()
#> $`2020-03-11`
#> $`2020-03-11`$home
#> 00-04 05-09 10-14 15-19 20-24 25-29 30-34 35-39 40-44 45-49 50-54 55-59 60-64
#> 1 1 1 1 1 1 1 1 1 1 1 1 1
#> 65-69 70-74 75+
#> 1 1 1
#>
#> $`2020-03-11`$work
#> 00-04 05-09 10-14 15-19 20-24 25-29 30-34 35-39 40-44 45-49 50-54 55-59 60-64
#> 1 1 1 1 1 1 1 1 1 1 1 1 1
#> 65-69 70-74 75+
#> 1 1 1
#>
#> $`2020-03-11`$school
#> 00-04 05-09 10-14 15-19 20-24 25-29 30-34 35-39 40-44 45-49 50-54 55-59 60-64
#> 1 1 1 1 1 1 1 1 1 1 1 1 1
#> 65-69 70-74 75+
#> 1 1 1
#>
#> $`2020-03-11`$other
#> 00-04 05-09 10-14 15-19 20-24 25-29 30-34 35-39 40-44 45-49 50-54 55-59 60-64
#> 1 1 1 1 1 1 1 1 1 1 1 1 1
#> 65-69 70-74 75+
#> 1 1 1The degree of openness can be queried similarly to the contact matrices:
Inspecting the module
As we saw above, we can use the $describe() method to
get a human readable summary of the module
act$describe()
#> # DiseasyActivity ############################################
#> Scenario: Overview
#> 2020-03-11
#> baseline 1
#>
#> Contact basis: Contact matrices for Denmark from the `contactdata` package and population data for
#> Denmark from the US Census Bureau.We can also inspect the individual elements of the module more
carefully such as the $scenario_matrix and
$risk_matrix. Both provide the internal representation of
the loaded scenario.
act$scenario_matrix
#> 2020-03-11
#> baseline 1
act$risk_matrix
#> 2020-03-11
#> home 1
#> work 1
#> school 1
#> other 1