,
Nick Golding [aut] ,
Aarathy Babu [aut] ,
Michael Lydeamore [aut] ,
Commonwealth of Australia AEC [cph],
Australian Bureau of Statistics ABS [cph]| Title: | Builds contact matrices using GAMs and population data |
|---|---|
| Description: | Builds contact matrices using GAMs and population data. This package incorporates data that is copyright Commonwealth of Australia (Australian Electoral Commission and Australian Bureau of Statistics) 2020. |
| Authors: | Nicholas Tierney [aut, cre] ,
Nick Golding [aut] ,
Aarathy Babu [aut] ,
Michael Lydeamore [aut] ,
Commonwealth of Australia AEC [cph],
Australian Bureau of Statistics ABS [cph] |
| Maintainer: | Nicholas Tierney <[email protected]> |
| License: | MIT + file LICENSE |
| Version: | 0.0.2.9000 |
| Built: | 2024-10-25 03:40:40 UTC |
| Source: | https://github.com/idem-lab/conmat |
Given a full name (Title Case) of an Australian State or Territory, produces the abbreviated state name.
abs_abbreviate_states(state_names)abs_abbreviate_states(state_names)
state_names |
vector of state names in long form |
shortened state names
abs_abbreviate_states("Victoria") abs_abbreviate_states(c("Victoria", "Queensland"))abs_abbreviate_states("Victoria") abs_abbreviate_states(c("Victoria", "Queensland"))
Return Australian Bureau of Statistics (ABS) age population data for a given Local Government Area (LGA) or state
abs_age_lga(lga_name) abs_age_state(state_name)abs_age_lga(lga_name) abs_age_state(state_name)
lga_name |
lga name - can be a partial match, e.g., although the official name might be "Albury (C)", "Albury" is fine. |
state_name |
shortened state name |
a conmat_population dataset containing: lga (or state),
lower.age.limit, year, and population.
abs_age_lga(c("Albury (C)", "Fairfield (C)")) abs_age_state(c("NSW", "VIC"))abs_age_lga(c("Albury (C)", "Fairfield (C)")) abs_age_state(c("NSW", "VIC"))
Return data on educated population for a given age and state or lga of Australia.
abs_age_education_state(state = NULL, age = NULL) abs_age_education_lga(lga = NULL, age = NULL)abs_age_education_state(state = NULL, age = NULL) abs_age_education_lga(lga = NULL, age = NULL)
state |
target Australian state name or a vector with multiple state names in its abbreviated form, such as "QLD", "NSW", or "TAS" |
age |
a numeric or numeric vector denoting ages between 0 to 115. The default is to return all ages. |
lga |
target Australian local government area (LGA) name, such as
"Fairfield (C)" or a vector with multiple lga names. See
|
dataset with information on the number of educated people belonging to a particular age, its total population and the corresponding proportion.
abs_age_education_state(state = "VIC") abs_age_education_state(state = "WA", age = 1:5) abs_age_education_state(state = c("QLD", "TAS"), age = 5) abs_age_education_lga(lga = c("Albury (C)", "Barcoo (S)"), age = 10)abs_age_education_state(state = "VIC") abs_age_education_state(state = "WA", age = 1:5) abs_age_education_state(state = c("QLD", "TAS"), age = 5) abs_age_education_lga(lga = c("Albury (C)", "Barcoo (S)"), age = 10)
Return data on employed population for a given age and state or lga of Australia
abs_age_work_lga(lga = NULL, age = NULL) abs_age_work_state(state = NULL, age = NULL)abs_age_work_lga(lga = NULL, age = NULL) abs_age_work_state(state = NULL, age = NULL)
lga |
target Australian local government area (LGA) name, such as
"Fairfield (C)" or a vector with multiple lga names. See
|
age |
a numeric or numeric vector denoting ages between 0 to 115. The default is to return all ages. |
state |
target Australian state name or a vector with multiple state names in its abbreviated form, such as "QLD", "NSW", or "TAS" |
data set with information on the number of employed people belonging to a particular age, its total population and the corresponding proportion.
abs_age_work_state(state = "NSW") abs_age_work_state(state = c("QLD", "TAS"), age = 5) abs_age_work_lga(lga = "Albany (C)", age = 1:5) abs_age_work_lga(lga = c("Albury (C)", "Barcoo (S)"), age = 39)abs_age_work_state(state = "NSW") abs_age_work_state(state = c("QLD", "TAS"), age = 5) abs_age_work_lga(lga = "Albany (C)", age = 1:5) abs_age_work_lga(lga = c("Albury (C)", "Barcoo (S)"), age = 39)
An internal dataset containing Australian Bureau of Statistics education data for
each age in 2016. The data is averaged across each state to provide an
overall average, and is used to provide estimated education populations for
model fitting in add_school_work_participation(), which is used in fit_single_contact_model(). The data is summarised from data_abs_state_education,
see ?data_abs_state_education for more details.
abs_avg_schoolabs_avg_school
A data frame with 116 rows and 2 variables:
0 to 115
fraction of population at school
Census of Population and Housing, 2016, TableBuilder
An internal dataset containing Australian Bureau of Statistics work data for
each age in 2016. The data is averaged across each state to provide an
overall average, and is used to provide estimated work populations for
model fitting in add_school_work_participation(), which is used in fit_single_contact_model(). The data is summarised from data_abs_state_work,
see ?data_abs_state_work for more details.
abs_avg_workabs_avg_work
A data frame with 116 rows and 2 variables:
0 to 115
fraction of population working.
Census of Population and Housing, 2016, TableBuilder
A dataset containing Australian Bureau of Statistics education data by state for 2006 to 2020
abs_education_stateabs_education_state
A data frame with 4194 rows and 5 variables:
year - 2020
state - short state or territory name
"Aboriginal and Torres Strait Islander" or "Non-Indigenous"
4 through to 21. Note that "4" is 4 or younger and "21" is actually 21+ (21 or older)
number of people full and part time
https://www.abs.gov.au/statistics/people/education/schools/2020#data-download (table 42B)
A dataset containing Australian Bureau of Statistics education data by state for 2020. These were interpolated into 1 year age bins. There are still some issued with the methods used, as the interpolated values are sometimes higher than the population.
abs_education_state_2020abs_education_state_2020
A data frame with 808 rows and 6 variables:
year - 2020
state - short state or territory name
0 to 100
number of people full and part time
"Government" or "Non-government"
population / population_interpolated
https://www.abs.gov.au/statistics/people/education/schools/2020#data-download (table 42B)
A dataset containing Australian Bureau of Statistics employment data by state for 2016
abs_employ_age_lgaabs_employ_age_lga
A data frame with 5600 rows and 8 variables:
year - 2016
state - short state or territory name
local government area name
age groups are as follows: 15-19, 20-24, 25-34, 35-44, 45-54, 55-64, 65-74, 75-84, 85+, total
total number of people employed
total number of people unemployed
total number of people in the labour force
sum of these totals...or thereabouts??
still need to finalise these columns
ABS.stat https://stat.data.abs.gov.au/Index.aspx? LABOUR > Employment and Unemployment > Labour force status > Census 2016, G43 labour status by age and sex (LGA)
A dataset containing Australian Bureau of Statistics household data for 2016. The data is filtered to "Total Households". Contains information on the number of people typically in a residence in the region and the number of households associated with those number of residents. This data is typically used to obtain the household size distributions to compute the per capita household size of a particular region.
abs_household_lgaabs_household_lga
A data frame with 4986 rows and 6 variables:
year - 2016
state - long state or territory name
name of LGA
Number of people typically in residence
number of households with that number of people
still need to clean this
https://stat.data.abs.gov.au/Index.aspx?Datasetcode=ABS_FAMILY_PROJ (downloaded the CSV) PEOPLE > People and Communities > Household Composition > Census 2016, T23 Household Composition By Number Of Persons Usually Resident (LGA)
A dataset containing Australian Bureau of Statistics official short state names, lga_code, and lga name.
abs_lga_lookupabs_lga_lookup
A data frame with 544 rows and 3 variables, arrange by state then LGA
state - short state or territory name
official lga code
lga name
A dataset containing Australian Bureau of Statistics population data by local government area (LGA) for age for 2016
abs_pop_age_lga_2016abs_pop_age_lga_2016
A data frame with 9918 rows and 6 variables:
year - 2020
state - short state or territory name
LGA name
age age band, 0-4, in 5 year increments up to 80-84, then 85+
number of people in a given lga in an age band
https://www.abs.gov.au/statistics/people/population/regional-population-age-and-sex
A dataset containing Australian Bureau of Statistics population data by local government area (LGA) for age for 2020
abs_pop_age_lga_2020abs_pop_age_lga_2020
A data frame with 9900 rows and 6 variables:
year - 2020
state - long state or territory name
LGA name
age group band, 0-4, in 5 year increments up to 80-84, then 85+
number of people in a given lga in an age band
https://www.abs.gov.au/statistics/people/population/regional-population-age-and-sex
Dataset containing Australian Bureau of Statistics state level population data for 2020
abs_state_ageabs_state_age
A data frame with 168 rows and 3 variables:
state - short state or territory name
age group in five year bins from 0 to 99, then 100+
population size
Un-abbreviate Australian state names
abs_unabbreviate_states(state_names)abs_unabbreviate_states(state_names)
state_names |
vector of state names in short form |
Longer state names
abs_unabbreviate_states("VIC") abs_unabbreviate_states(c("VIC", "QLD"))abs_unabbreviate_states("VIC") abs_unabbreviate_states(c("VIC", "QLD"))
For modelling purposes it is useful to have a feature that is the absolute
difference between age_from and age_to columns.
add_intergenerational(data)add_intergenerational(data)
data |
data.frame with columns |
data.frame with extra column, intergenerational
polymod_contact <- get_polymod_contact_data() polymod_contact %>% add_intergenerational()polymod_contact <- get_polymod_contact_data() polymod_contact %>% add_intergenerational()
This function adds three main groups of features to the data. It is used
internally in fit_single_contact_model() and predict_contacts_1y().
It requires columns named age_to and age_from. The three types of
features it adds are described below:
Population distribution of contact ages from the function
add_population_age_to(), which requires a column called "age_to"
representing the age of the person who had contact. It creates a column
called pop_age_to. add_population_age_to() takes an extra argument
for population, which defaults to get_polymod_population(), but needs
to be a conmat_population object, which specifies the age and
population characteristics, or a data frame with columns,
lower.age.limit, and population.
School work participation, which is from the function
add_school_work_participation(). This requires columns age_to and
age_from, but will operate on any column starting with age and adds
columns: school_probability, work_probability,
school_year_probability, and school_weighted_pop_fraction.
Offset is added on to the data using add_offset(). This requires
variables school_weighted_pop_fraction (from
add_school_work_participation()) and pop_age_to (from
add_school_work_participation()). It adds two columns,
log_contactable_population_school, and log_contactable_population.
add_modelling_features( contact_data, school_demographics = NULL, work_demographics = NULL, population = get_polymod_population() )add_modelling_features( contact_data, school_demographics = NULL, work_demographics = NULL, population = get_polymod_population() )
contact_data |
contact data with columns |
school_demographics |
(optional) defaults to census average proportion
at school. You can provide a dataset with columns, "age" (numeric), and
"school_fraction" (0-1), if you would like to specify these
details. See |
work_demographics |
(optional) defaults to census average proportion
employed. You can provide a dataset with columns, "age" (numeric), and
"work_fraction", if you would like to specify these details. See
|
population |
the |
data frame with 11 extra columns - the contents of contact_data,
plus: pop_age_to, school_fraction_age_from, work_fraction_age_from,
school_fraction_age_to, work_fraction_age_to, school_probability,
work_probability, school_year_probability, school_weighted_pop_fraction,
log_contactable_population_school, and log_contactable_population.
age_min <- 10 age_max <- 15 all_ages <- age_min:age_max library(tidyr) example_df <- expand_grid( age_from = all_ages, age_to = all_ages, ) add_modelling_features(example_df) add_modelling_features( example_df, school_demographics = conmat_original_school_demographics, work_demographics = conmat_original_work_demographics )age_min <- 10 age_max <- 15 all_ages <- age_min:age_max library(tidyr) example_df <- expand_grid( age_from = all_ages, age_to = all_ages, ) add_modelling_features(example_df) add_modelling_features( example_df, school_demographics = conmat_original_school_demographics, work_demographics = conmat_original_work_demographics )
Mostly used internally in add_modelling_features(). Adds two
offset variables to be used in fit_single_contact_model():
log_contactable_population_school, and
log_contactable_population.
These two variables require variables school_weighted_pop_fraction (from
add_school_work_participation()) and pop_age_to (from
add_school_work_participation()). This provides separate offsets
for school setting when compared to the other settings such as home,
work and other. The offset for school captures cohorting of students
for schools and takes the logarithm of the weighted combination of
contact population age distribution & school year probability calculated
in add_school_work_participation(). See "details" for more information.
add_offset(contact_data)add_offset(contact_data)
contact_data |
contact data - must contain columns |
why double offsets? There are two offsets specified, once in the
model formula, and once in the "offset" argument of mgcv::bam. The
offsets get added together when the model first fit. In addition, the
setting specific offset from offset_variable, which is included in the
GAM model as ... + offset(log_contactable_population) is used in
prediction, whereas the other offset, included as an argument in the GAM
as offset = log(participants) is only included when the model is
initially created. See more detail in fit_single_contact_model().
data.frame of contact_data with two extra columns:
log_contactable_population_school and log_contactable_population
Nick Golding
age_min <- 10 age_max <- 15 all_ages <- age_min:age_max library(tidyr) example_df <- expand_grid( age_from = all_ages, age_to = all_ages, ) example_df %>% add_population_age_to() %>% add_school_work_participation() %>% add_offset()age_min <- 10 age_max <- 15 all_ages <- age_min:age_max library(tidyr) example_df <- expand_grid( age_from = all_ages, age_to = all_ages, ) example_df %>% add_population_age_to() %>% add_school_work_participation() %>% add_offset()
Adds the population distribution of contact ages. Requires a column called
"age_to", representing the contact age - the age of the person who had
contact. It creates a column, pop_age_to. The population argument
defaults to get_polymod_population(), which is a conmat_population
object, which has age and population specified. But this can also be
a data frame with columns, lower.age.limit, and population. If
population is 'polymod' then use the participant-weighted average of
POLYMOD country/year distributions. It adds the population via
interpolation, using get_age_population_function() to create a
function that generates population from ages.
add_population_age_to(contact_data, population = get_polymod_population())add_population_age_to(contact_data, population = get_polymod_population())
contact_data |
contact data containing columns |
population |
Defaults to |
data frame
age_min <- 10 age_max <- 15 all_ages <- age_min:age_max library(tidyr) example_df <- expand_grid( age_from = all_ages, age_to = all_ages, ) add_population_age_to(example_df)age_min <- 10 age_max <- 15 all_ages <- age_min:age_max library(tidyr) example_df <- expand_grid( age_from = all_ages, age_to = all_ages, ) add_population_age_to(example_df)
Add fractions of the population in each age group that attend school/work
(average FTE) to compute the probability that both participant and
contact attend school/work. Requires columns age_to and age_from.
Note that it will operate on any column starting with age. Adds columns:
school_probability, work_probability, school_year_probability, and
school_weighted_pop_fraction. The columns school_probability and
work_probability represent the probability a person of the other age
goes to the same work/school. school_year_probability represents the
probability that a person of the other age would be in the same school
year. school_weighted_pop_fraction represents the weighted combination
of contact population age distribution & school year probability, so that
if the contact is in the same school year, the weight is 1, and otherwise
it is the population age fraction. This can be used as an offset, so that
population age distribution can be used outside the classroom, but does
not affect classroom contacts (which due to cohorting and regularised
class sizes are unlikely to depend on the population age distribution).
add_school_work_participation( contact_data, school_demographics = NULL, work_demographics = NULL )add_school_work_participation( contact_data, school_demographics = NULL, work_demographics = NULL )
contact_data |
contact data containing columns: |
school_demographics |
(optional) defaults to census average proportion
at school. You can provide a dataset with columns, "age" (numeric), and
"school_fraction" (0-1), if you would like to specify these
details. See |
work_demographics |
(optional) defaults to census average proportion
employed. You can provide a dataset with columns, "age" (numeric), and
"work_fraction", if you would like to specify these details. See
|
dataset with 9 extra columns: school_fraction_age_from, work_fraction_age_from, school_fraction_age_to, work_fraction_age_to, school_probability, work_probability, school_year_probability, and school_weighted_pop_fraction.
To use previous approach input the arguments school_demographics and
work_demographics with conmat_original_school_demographics and conmat_original_work_demographics, respectively.
age_min <- 10 age_max <- 15 all_ages <- age_min:age_max library(tidyr) example_df <- expand_grid( age_from = all_ages, age_to = all_ages, ) example_df %>% add_population_age_to() %>% add_school_work_participation() example_df %>% add_population_age_to() %>% add_school_work_participation( school_demographics = conmat_original_school_demographics, work_demographics = conmat_original_work_demographics )age_min <- 10 age_max <- 15 all_ages <- age_min:age_max library(tidyr) example_df <- expand_grid( age_from = all_ages, age_to = all_ages, ) example_df %>% add_population_age_to() %>% add_school_work_participation() example_df %>% add_population_age_to() %>% add_school_work_participation( school_demographics = conmat_original_school_demographics, work_demographics = conmat_original_work_demographics )
This function adds 6 columns to assist with describing various age based interactions for model fitting. Requires that the age columns are called "age_from", and "age_to"
add_symmetrical_features(data)add_symmetrical_features(data)
data |
data.frame with columns, |
data.frame with 6 more columns, gam_age_offdiag, gam_age_offdiag_2, gam_age_diag_prod, gam_age_diag_sum, gam_age_pmax, gam_age_pmin,
vec_age <- 0:2 dat_age <- expand.grid( age_from = vec_age, age_to = vec_age ) add_symmetrical_features(dat_age)vec_age <- 0:2 dat_age <- expand.grid( age_from = vec_age, age_to = vec_age ) add_symmetrical_features(dat_age)
Accessing conmat attributes
age(x) age_label(x) ## Default S3 method: age_label(x) ## S3 method for class 'conmat_population' age_label(x) population_label(x) ## Default S3 method: population_label(x) ## S3 method for class 'conmat_population' population_label(x) population(x)age(x) age_label(x) ## Default S3 method: age_label(x) ## S3 method for class 'conmat_population' age_label(x) population_label(x) ## Default S3 method: population_label(x) ## S3 method for class 'conmat_population' population_label(x) population(x)
x |
conmat_population data frame |
age or population symbol or label
## Not run: perth <- abs_age_lga("Perth (C)") age(perth) age_label(perth) population(perth) population_label(perth) ## End(Not run)## Not run: perth <- abs_age_lga("Perth (C)") age(perth) age_label(perth) population(perth) population_label(perth) ## End(Not run)
Extract age break attribute information
age_breaks(x) ## S3 method for class 'conmat_age_matrix' age_breaks(x) ## S3 method for class 'conmat_setting_prediction_matrix' age_breaks(x) ## S3 method for class 'setting_data' age_breaks(x) ## S3 method for class 'ngm_setting_matrix' age_breaks(x) ## S3 method for class 'setting_vaccination_matrix' age_breaks(x) ## S3 method for class 'numeric' age_breaks(x) ## S3 method for class 'matrix' age_breaks(x) ## S3 method for class 'array' age_breaks(x) ## S3 method for class 'predicted_contacts' age_breaks(x) ## S3 method for class 'transmission_probability_matrix' age_breaks(x) ## S3 method for class 'setting_contact_model' age_breaks(x) ## Default S3 method: age_breaks(x)age_breaks(x) ## S3 method for class 'conmat_age_matrix' age_breaks(x) ## S3 method for class 'conmat_setting_prediction_matrix' age_breaks(x) ## S3 method for class 'setting_data' age_breaks(x) ## S3 method for class 'ngm_setting_matrix' age_breaks(x) ## S3 method for class 'setting_vaccination_matrix' age_breaks(x) ## S3 method for class 'numeric' age_breaks(x) ## S3 method for class 'matrix' age_breaks(x) ## S3 method for class 'array' age_breaks(x) ## S3 method for class 'predicted_contacts' age_breaks(x) ## S3 method for class 'transmission_probability_matrix' age_breaks(x) ## S3 method for class 'setting_contact_model' age_breaks(x) ## Default S3 method: age_breaks(x)
x |
an object containing age break information |
age breaks character vector
age_breaks(conmat_age_matrix): Get age break information
age_breaks(conmat_setting_prediction_matrix): Get age break information
age_breaks(setting_data): Get age break information
age_breaks(ngm_setting_matrix): Get age break information
age_breaks(setting_vaccination_matrix): Get age break information
age_breaks(numeric): Get age break information
age_breaks(matrix): Get age break information
age_breaks(array): Get age break information
age_breaks(predicted_contacts): Get age break information
age_breaks(transmission_probability_matrix): Get age break information
age_breaks(setting_contact_model): Get age break information
age_breaks(default): Get age break information
age_breaks <- c(0, 5, 19, 15) age_break_names <- c("[0,5)", "[5,10)", "[10, 15)") age_mat <- matrix( runif(9), nrow = 3, ncol = 3, dimnames = list( age_break_names, age_break_names ) ) age_mat <- new_age_matrix(age_mat, age_breaks) age_breaks(age_mat)age_breaks <- c(0, 5, 19, 15) age_break_names <- c("[0,5)", "[5,10)", "[10, 15)") age_mat <- matrix( runif(9), nrow = 3, ncol = 3, dimnames = list( age_break_names, age_break_names ) ) age_mat <- new_age_matrix(age_mat, age_breaks) age_breaks(age_mat)
A dataset containing age lower and upper levels with age group
age_group_lookupage_group_lookup
A data frame with 21 rows and 3 variables:
Lower age
upper age
age group as a factor
This function helps clean up datasets of population data, which
might be similar to socialmixr::wpp_age() or a dataset with columns
representing: population, location, age, and year. If age is numeric, it
groups ages into age groups with 5 year bins (0-4, 5-9, etc). It then
separates age groups into two column of these lower and upper limits.
Finally, it filters data passed to the specified year and location. If no
year or location is provided then all years or locations are used.
age_population( data, location_col = NULL, location = NULL, age_col, year_col = NULL, year = NULL )age_population( data, location_col = NULL, location = NULL, age_col, year_col = NULL, year = NULL )
data |
dataset containing information on population for a given age, country, and year |
location_col |
bare variable name for the column with location
information. If using, both |
location |
character vector with location names. If using, both
|
age_col |
bare variable name for the column with age information |
year_col |
bare variable name for the column with year information. If
using, both |
year |
numeric vector representing the desired year(s). If using, both
|
tidy dataset with information on population of different age bands
world_data <- socialmixr::wpp_age() world_data # Tidy data for multiple locations across different years age_population( data = world_data, location_col = country, location = c("Asia", "Afghanistan"), age_col = lower.age.limit, year_col = year, year = c(2010:2020) ) # Tidy data for a given location irrespective of year age_population( data = world_data, location_col = country, location = "Afghanistan", age_col = lower.age.limit ) # Tidy data for a given location irrespective of location age_population( data = world_data, age_col = lower.age.limit ) age_population( data = world_data, age_col = lower.age.limit, year_col = year, year = c(2011:2015) ) # Tidy datasets with age groups population_age_groups <- abs_pop_age_lga_2020 population_age_groups age_population( data = population_age_groups, age_col = age_group, year_col = year, year = 2020 ) # Tidy datasets with numeric age population_numeric_age <- abs_age_state("WA") population_numeric_age age_population( data = population_numeric_age, age_col = lower.age.limit, year_col = year, year = 2020 )world_data <- socialmixr::wpp_age() world_data # Tidy data for multiple locations across different years age_population( data = world_data, location_col = country, location = c("Asia", "Afghanistan"), age_col = lower.age.limit, year_col = year, year = c(2010:2020) ) # Tidy data for a given location irrespective of year age_population( data = world_data, location_col = country, location = "Afghanistan", age_col = lower.age.limit ) # Tidy data for a given location irrespective of location age_population( data = world_data, age_col = lower.age.limit ) age_population( data = world_data, age_col = lower.age.limit, year_col = year, year = c(2011:2015) ) # Tidy datasets with age groups population_age_groups <- abs_pop_age_lga_2020 population_age_groups age_population( data = population_age_groups, age_col = age_group, year_col = year, year = 2020 ) # Tidy datasets with numeric age population_numeric_age <- abs_age_state("WA") population_numeric_age age_population( data = population_numeric_age, age_col = lower.age.limit, year_col = year, year = 2020 )
Aggregates contacts rate from, say, a 1 year level into
provided age breaks, weighting the contact rate by the specified age
population. For example, if you specify breaks as c(0, 5, 10, 15, Inf),
it will return age groups as 0-5, 5-10, 10-15, and 15+ (Inf). Used
internally within predict_contacts(), although can be used by users.
aggregate_predicted_contacts( predicted_contacts_1y, population, age_breaks = c(seq(0, 75, by = 5), Inf) )aggregate_predicted_contacts( predicted_contacts_1y, population, age_breaks = c(seq(0, 75, by = 5), Inf) )
predicted_contacts_1y |
contacts in 1 year breaks (could technically
by in other year breaks). Data must contain columns, |
population |
a |
age_breaks |
vector of ages. Default: c(seq(0, 75, by = 5), Inf) |
data frame with columns, age_group_from, age_group_to, and
contacts, which is the aggregated model.
fairfield <- abs_age_lga("Fairfield (C)") fairfield # We can predict the contact rate for Fairfield from the existing contact # data, say, between the age groups of 0-15 in 5 year bins for school: fairfield_contacts_1 <- predict_contacts_1y( model = polymod_setting_models$home, population = fairfield, age_min = 0, age_max = 15 ) fairfield_contacts_1 aggregated_fairfield <- aggregate_predicted_contacts( predicted_contacts_1y = fairfield_contacts_1, population = fairfield, age_breaks = c(0, 5, 10, 15, Inf) ) aggregated_fairfieldfairfield <- abs_age_lga("Fairfield (C)") fairfield # We can predict the contact rate for Fairfield from the existing contact # data, say, between the age groups of 0-15 in 5 year bins for school: fairfield_contacts_1 <- predict_contacts_1y( model = polymod_setting_models$home, population = fairfield, age_min = 0, age_max = 15 ) fairfield_contacts_1 aggregated_fairfield <- aggregate_predicted_contacts( predicted_contacts_1y = fairfield_contacts_1, population = fairfield, age_breaks = c(0, 5, 10, 15, Inf) ) aggregated_fairfield
Applies the effect of vaccination on the next generation of infections, to understand and describe the reduction of acquisition and transmission in each age group.
apply_vaccination(ngm, data, coverage_col, acquisition_col, transmission_col)apply_vaccination(ngm, data, coverage_col, acquisition_col, transmission_col)
ngm |
next generation matrices. See |
data |
data frame with location specific information on vaccine coverage, efficacy of acquisition/susceptibility and efficacy of transmission/infectiousness for the ordered age groups from lowest to highest of the next generation matrix |
coverage_col |
bare variable name for the column with information on vaccine coverage by age groups |
acquisition_col |
bare variable name for the column with information on efficacy of acquisition |
transmission_col |
bare variable name for the column with information on efficacy of transmission |
Vaccination improves a person's immunity from a disease. When a sizeable section of the population receives vaccinations or when vaccine coverage is sufficient enough, the likelihood that the unvaccinated population will contract the disease is decreased. This helps to slow infectious disease spread as well as lessen its severity. For this reason, it is important to understand how much of a reduction in probability of acquisition (the likelihood that an individual will contract the disease), and probability of transmission (the likelihood that an individual will spread the disease after contracting it), has occurred as an the effect of vaccination, in other words the effect of vaccination on the next generation of infections.
apply_vaccination returns the percentage reduction in acquisition and
transmission in each age group. It does this by taking the outer product
of these reductions in acquisition and transmission by age group, creating
a transmission reduction matrix. The next generation matrices with the
vaccination effects applied are then produced using the obtained
transmission reduction matrix and the next generation matrices passed to
the function as an argument.
list of contact matrices, one for each setting with reduction in transmission matching the next generation matrices
# examples take 20 second to run so skipping ## Not run: # example data frame with vaccine coverage, acquisition and transmission # efficacy of different age groups vaccination_effect_example_data # Generate next generation matrices perth <- abs_age_lga("Perth (C)") perth_hh <- get_abs_per_capita_household_size(lga = "Perth (C)") age_breaks_0_80 <- c(seq(0, 80, by = 5), Inf) # refit the model - note that the default if age_breaks isn't specified is # 0 to 75 perth_contact_0_80 <- extrapolate_polymod( perth, per_capita_household_size = perth_hh, age_breaks = age_breaks_0_80 ) perth_ngm_0_80 <- generate_ngm(perth_contact_0_80, age_breaks = age_breaks_0_80, per_capita_household_size = perth_hh, R_target = 1.5 ) # In the old way we used to be able to pass age_breaks_0_80 along generate_ngm_oz( lga_name = "Perth (C)", age_breaks = age_breaks_0_80, R_target = 1.5 ) # another way to do this using the previous method for generating NGMs # The number of age breaks must match the vaccination effect data ngm_nsw <- generate_ngm_oz( state_name = "NSW", age_breaks = c(seq(0, 80, by = 5), Inf), R_target = 1.5 ) # Apply vaccination effect to next generation matrices ngm_nsw_vacc <- apply_vaccination( ngm = ngm_nsw, data = vaccination_effect_example_data, coverage_col = coverage, acquisition_col = acquisition, transmission_col = transmission ) ## End(Not run)# examples take 20 second to run so skipping ## Not run: # example data frame with vaccine coverage, acquisition and transmission # efficacy of different age groups vaccination_effect_example_data # Generate next generation matrices perth <- abs_age_lga("Perth (C)") perth_hh <- get_abs_per_capita_household_size(lga = "Perth (C)") age_breaks_0_80 <- c(seq(0, 80, by = 5), Inf) # refit the model - note that the default if age_breaks isn't specified is # 0 to 75 perth_contact_0_80 <- extrapolate_polymod( perth, per_capita_household_size = perth_hh, age_breaks = age_breaks_0_80 ) perth_ngm_0_80 <- generate_ngm(perth_contact_0_80, age_breaks = age_breaks_0_80, per_capita_household_size = perth_hh, R_target = 1.5 ) # In the old way we used to be able to pass age_breaks_0_80 along generate_ngm_oz( lga_name = "Perth (C)", age_breaks = age_breaks_0_80, R_target = 1.5 ) # another way to do this using the previous method for generating NGMs # The number of age breaks must match the vaccination effect data ngm_nsw <- generate_ngm_oz( state_name = "NSW", age_breaks = c(seq(0, 80, by = 5), Inf), R_target = 1.5 ) # Apply vaccination effect to next generation matrices ngm_nsw_vacc <- apply_vaccination( ngm = ngm_nsw, data = vaccination_effect_example_data, coverage_col = coverage, acquisition_col = acquisition, transmission_col = transmission ) ## End(Not run)
Convert to conmat population
as_conmat_population(data, ...) ## Default S3 method: as_conmat_population(data, ...) ## S3 method for class 'data.frame' as_conmat_population(data, age, population, ...) ## S3 method for class 'list' as_conmat_population(data, age, population, ...) ## S3 method for class 'grouped_df' as_conmat_population(data, age, population, ...)as_conmat_population(data, ...) ## Default S3 method: as_conmat_population(data, ...) ## S3 method for class 'data.frame' as_conmat_population(data, age, population, ...) ## S3 method for class 'list' as_conmat_population(data, age, population, ...) ## S3 method for class 'grouped_df' as_conmat_population(data, age, population, ...)
data |
data.frame |
... |
extra arguments |
age |
age column - an unquoted variable of numeric integer ages |
population |
population column - an unquoted variable, numeric value |
some_age_pop <- data.frame( age = 1:10, pop = 101:110 ) some_age_pop as_conmat_population( some_age_pop, age = age, population = pop )some_age_pop <- data.frame( age = 1:10, pop = 101:110 ) some_age_pop as_conmat_population( some_age_pop, age = age, population = pop )
This will also calculate an all matrix, if all is not specified. This
is the sum of all other matrices.
as_setting_prediction_matrix(list_matrix, age_breaks, ...)as_setting_prediction_matrix(list_matrix, age_breaks, ...)
list_matrix |
list of matrices |
age_breaks |
numeric vector of ages |
... |
extra arguments (currently not used) |
object of class setting prediction matrix
age_breaks_0_80_plus <- c(seq(0, 80, by = 10), Inf) one_by_nine <- matrix(1, nrow = 9, ncol = 9) mat_list <- list( home = one_by_nine, work = one_by_nine ) mat_list mat_set <- as_setting_prediction_matrix( mat_list, age_breaks = age_breaks_0_80_plus ) mat_setage_breaks_0_80_plus <- c(seq(0, 80, by = 10), Inf) one_by_nine <- matrix(1, nrow = 9, ncol = 9) mat_list <- list( home = one_by_nine, work = one_by_nine ) mat_list mat_set <- as_setting_prediction_matrix( mat_list, age_breaks = age_breaks_0_80_plus ) mat_set
Plot setting matrices using ggplot2
## S3 method for class 'conmat_age_matrix' autoplot(object, ..., title = "Contact Matrices") ## S3 method for class 'conmat_setting_prediction_matrix' autoplot(object, ..., title = "Setting-specific synthetic contact matrices") ## S3 method for class 'transmission_probability_matrix' autoplot( object, ..., title = "Setting-specific transmission probability matrices" ) ## S3 method for class 'ngm_setting_matrix' autoplot(object, ..., title = "Setting-specific NGM matrices") ## S3 method for class 'setting_vaccination_matrix' autoplot(object, ..., title = "Setting-specific vaccination matrices")## S3 method for class 'conmat_age_matrix' autoplot(object, ..., title = "Contact Matrices") ## S3 method for class 'conmat_setting_prediction_matrix' autoplot(object, ..., title = "Setting-specific synthetic contact matrices") ## S3 method for class 'transmission_probability_matrix' autoplot( object, ..., title = "Setting-specific transmission probability matrices" ) ## S3 method for class 'ngm_setting_matrix' autoplot(object, ..., title = "Setting-specific NGM matrices") ## S3 method for class 'setting_vaccination_matrix' autoplot(object, ..., title = "Setting-specific vaccination matrices")
object |
A matrix or a list of square matrices, with row and column names indicating the age groups. |
... |
Other arguments passed on |
title |
Title to give to plot setting matrices. Defaults are provided for certain objects |
a ggplot visualisation of contact rates
## Not run: if (interactive()) { polymod_contact_data <- get_polymod_setting_data() polymod_survey_data <- get_polymod_population() setting_models <- fit_setting_contacts( contact_data_list = polymod_contact_data, population = polymod_survey_data ) fairfield <- abs_age_lga("Fairfield (C)") fairfield_hh_size <- get_abs_per_capita_household_size(lga = "Fairfield (C)") synthetic_settings_5y_fairfield_hh <- predict_setting_contacts( population = fairfield, contact_model = setting_models, age_breaks = c(seq(0, 85, by = 5), Inf), per_capita_household_size = fairfield_hh_size ) # Plotting synthetic contact matrices across all settings autoplot( object = synthetic_settings_5y_fairfield_hh, title = "Setting specific synthetic contact matrices" ) # Work setting specific synthetic contact matrices autoplot( object = synthetic_settings_5y_fairfield_hh$work, title = "Work" ) } ## End(Not run)## Not run: if (interactive()) { polymod_contact_data <- get_polymod_setting_data() polymod_survey_data <- get_polymod_population() setting_models <- fit_setting_contacts( contact_data_list = polymod_contact_data, population = polymod_survey_data ) fairfield <- abs_age_lga("Fairfield (C)") fairfield_hh_size <- get_abs_per_capita_household_size(lga = "Fairfield (C)") synthetic_settings_5y_fairfield_hh <- predict_setting_contacts( population = fairfield, contact_model = setting_models, age_breaks = c(seq(0, 85, by = 5), Inf), per_capita_household_size = fairfield_hh_size ) # Plotting synthetic contact matrices across all settings autoplot( object = synthetic_settings_5y_fairfield_hh, title = "Setting specific synthetic contact matrices" ) # Work setting specific synthetic contact matrices autoplot( object = synthetic_settings_5y_fairfield_hh$work, title = "Work" ) } ## End(Not run)
An internal dataset containing the original estimates of which fraction of
ages were attending school in Australia. These can be used inside of
fit_single_contact_model() and fit_setting_contacts().
conmat_original_school_demographicsconmat_original_school_demographics
A data frame with 121 rows and 2 variables:
0 to 120
fraction of population at school
Census of Population and Housing, 2016, TableBuilder
An internal dataset containing the original estimates of which fraction of
ages were working in Australia. These can be used inside of
fit_single_contact_model() and fit_setting_contacts().
conmat_original_work_demographicsconmat_original_work_demographics
A data frame with 121 rows and 2 variables:
0 to 120
fraction of population working.
Census of Population and Housing, 2016, TableBuilder
A conmat population is a dataframe that stores which columns represent the age and population information. This is useful as it means we can refer to this information throughout other functions in the conmat package without needing to specify or hard code which columns represent the age and population information.
conmat_population(data, age, population)conmat_population(data, age, population)
data |
data.frame |
age |
bare name representing the age column |
population |
bare name representing the population column |
a data frame with age and population attributes
perth <- abs_age_lga("Perth (C)")perth <- abs_age_lga("Perth (C)")
A dataset containing Australian Bureau of Statistics education data by
lga for 2016. The data sourced from 2016 Census - Employment, Income and
Education through TableBuilder have been randomly adjusted by the ABS to
avoid the release of confidential data. As a result of this, there are
some cases where the estimated number of people being educated is higher
than the population of those people. Such cases have been flagged under the
anomaly_flag variable.
data_abs_lga_educationdata_abs_lga_education
A data frame with 64,264 rows and 8 variables:
2016, data is based on 2016 Census of Population and Housing.
String denoting abbreviated name of state or territory, for example, 'NSW', 'VIC', and 'QLD'.
String denoting the official name of Local Government Area. For example, 'Albury (C).'
Ages from 0 to 115.
Number of people educated including students with full-time, part-time status, as well as the people who mentioned just the type of educational institution they attend and not their student status.
Number depicting the total population belonging to the age.
Number denoting the measure of the ratio of educated population and total population belonging to the age i.e, population_educated / total_population
Logical variable flagging abnormal observations. E.g., total population lesser than population_educated as TRUE.
Census of Population and Housing, 2016, TableBuilder
A dataset containing Australian Bureau of Statistics labour force population
data by lga for 2016. The data sourced from 2016 Census - Employment,
Income and Education through TableBuilder have been randomly adjusted by
the ABS to avoid the release of confidential data. As a result of this,
there are some cases where the estimated number of people being employed
is higher than the population of those people. Such cases have been
flagged under the anomaly_flag variable.
data_abs_lga_workdata_abs_lga_work
A data frame with 64,496 rows and 8 variables:
2016, as data is from 2016 Census of Population and Housing.
String denoting the abbreviated name of state or territory name such as 'NSW', 'VIC', 'QLD' etc.
String denoting the official name of Local Government Area. For example, 'Albury (C).'
Ages from 0 to 115.
Number of people employed including people with full-time, part-time employment status.
Total population of age in row.
The ratio of employed population and total population belonging to the age i.e, employed_population/ total_population.
Logical variable flagging abnormal observations, such as total population lesser than employed_population as TRUE.
Census of Population and Housing, 2016, TableBuilder
A dataset containing Australian Bureau of Statistics education data by state for 2016. The data sourced from 2016 Census - Employment, Income and Education through TableBuilder have been randomly adjusted by the ABS to avoid the release of confidential data.
data_abs_state_educationdata_abs_state_education
A data frame with 1044 rows and 6 variables:
2016, as data is from 2016 Census of Population and Housing.
String of abbreviated name of state or territory names, e.g., 'NSW', 'VIC', 'QLD' and so on.
Ages from 0 to 115.
Number of people educated, including students with full-time, part-time status, and people who mentioned only the type of educational institution they attend and not their student status.
Total population belonging to age in a row.
The ratio of educated population and total population belonging to the age i.e, population_educated / total_population
Census of Population and Housing, 2016, TableBuilder
A dataset containing Australian Bureau of Statistics labour force population data by state for 2016. The data sourced from 2016 Census - Employment, Income and Education through TableBuilder have been randomly adjusted by the ABS to avoid the release of confidential data.
data_abs_state_workdata_abs_state_work
A data frame with 1044 rows and 6 variables:
2016, as data is from 2016 Census of Population and Housing.
String. Abbreviated name of state or territory, e.g., 'NSW', 'VIC', 'QLD' and so on.
Ages from 0 to 115.
Number of people employed including people with full-time, part-time employment status.
Total population belonging to the age.
The ratio of employed population and total population belonging to the age i.e, employed_population/ total_population
Census of Population and Housing, 2016, TableBuilder
A dataset containing data from https://www.nature.com/articles/s41591-020-0962-9#code-availability When using this data, ensure that you cite the original authors at:
davies_age_extendeddavies_age_extended
A data frame of the probability of transmission from a case to a contact. There are 101 rows and 4 variables.
from 0 to 100
Estimate of fraction with clinical symptoms, or the age-specific proportion of infections resulting in clinical symptoms inferred by applying a smoothing spline to the mean estimates from Davies et al.
Age specific parameters of the relative susceptibility to infection inferred from a smoothing-spline estimate of the mean relative susceptibility estimate from Davies et al.
Re-estimated parameter of the susceptibility profile for under-16s that is estimated in a similar way but to the age-distribution of infections in England from the UK ONS prevalence survey rather than case counts which may undercount children
"Davies, N.G., Klepac, P., Liu, Y. et al. Age-dependent effects in the transmission and control of COVID-19 epidemics. Nat Med 26, 1205–1211 (2020). https://doi.org/10.1038/s41591-020-0962-9"
Given a named list of contact datasets (with names giving
the setting, and assumed to together make up the full set of contacts for
individuals in the survey), a representative population distribution for
the survey, and a set of age breaks at which to aggregate contacts, return
a set of predicted contact matrices for each setting, and for all combined.
Note that this function is parallelisable with future, and will be
impacted by any future plans provided.
estimate_setting_contacts( contact_data_list, survey_population, prediction_population = survey_population, age_breaks, per_capita_household_size = NULL, symmetrical = TRUE, school_demographics = NULL, work_demographics = NULL )estimate_setting_contacts( contact_data_list, survey_population, prediction_population = survey_population, age_breaks, per_capita_household_size = NULL, symmetrical = TRUE, school_demographics = NULL, work_demographics = NULL )
contact_data_list |
list of data sets with information on the contacts of individuals at different settings |
survey_population |
representative population distribution for the survey |
prediction_population |
population for prediction. The default value set is survey_population |
age_breaks |
vector depicting age values. For example,
|
per_capita_household_size |
Optional (defaults to NULL). When set, it
adjusts the household contact matrix by some per capita household size.
To set it, provide a single number, the per capita household size. More
information is provided below in Details. See
|
symmetrical |
whether to enforce symmetrical terms in the model.
Defaults to TRUE. See |
school_demographics |
(optional) defaults to census average proportion
at school. You can provide a dataset with columns, "age" (numeric), and
"school_fraction" (0-1), if you would like to specify these
details. See |
work_demographics |
(optional) defaults to census average proportion
employed. You can provide a dataset with columns, "age" (numeric), and
"work_fraction", if you would like to specify these details. See
|
predicted setting specific contact matrices, and for all combined
## Not run: # takes a long time to run settings_estimated_contacts <- estimate_setting_contacts( contact_data_list = get_polymod_setting_data(), survey_population = get_polymod_population(), prediction_population = get_polymod_population(), age_breaks = c(seq(0, 85, by = 5), Inf), per_capita_household_size = NULL ) # or predict to fairfield fairfield_hh <- get_abs_per_capita_household_size(lga = "Fairfield (C)") contact_model_pred_est <- estimate_setting_contacts( contact_data_list = get_polymod_setting_data(), survey_population = get_polymod_population(), prediction_population = abs_age_lga("Fairfield (C)"), age_breaks = c(seq(0, 85, by = 5), Inf), per_capita_household_size = fairfield_hh ) # or use different populations in school or work demographics fairfield_hh <- get_abs_per_capita_household_size(lga = "Fairfield (C)") contact_model_pred_est <- estimate_setting_contacts( contact_data_list = get_polymod_setting_data(), survey_population = get_polymod_population(), prediction_population = abs_age_lga("Fairfield (C)"), age_breaks = c(seq(0, 85, by = 5), Inf), per_capita_household_size = fairfield_hh, school_demographics = conmat_original_school_demographics, work_demographics = conmat_original_work_demographics ) # or use non-symmetric model terms contact_model_pred_est <- estimate_setting_contacts( contact_data_list = get_polymod_setting_data(), survey_population = get_polymod_population(), prediction_population = abs_age_lga("Fairfield (C)"), age_breaks = c(seq(0, 85, by = 5), Inf), per_capita_household_size = fairfield_hh, symmetrical = FALSE ) ## End(Not run)## Not run: # takes a long time to run settings_estimated_contacts <- estimate_setting_contacts( contact_data_list = get_polymod_setting_data(), survey_population = get_polymod_population(), prediction_population = get_polymod_population(), age_breaks = c(seq(0, 85, by = 5), Inf), per_capita_household_size = NULL ) # or predict to fairfield fairfield_hh <- get_abs_per_capita_household_size(lga = "Fairfield (C)") contact_model_pred_est <- estimate_setting_contacts( contact_data_list = get_polymod_setting_data(), survey_population = get_polymod_population(), prediction_population = abs_age_lga("Fairfield (C)"), age_breaks = c(seq(0, 85, by = 5), Inf), per_capita_household_size = fairfield_hh ) # or use different populations in school or work demographics fairfield_hh <- get_abs_per_capita_household_size(lga = "Fairfield (C)") contact_model_pred_est <- estimate_setting_contacts( contact_data_list = get_polymod_setting_data(), survey_population = get_polymod_population(), prediction_population = abs_age_lga("Fairfield (C)"), age_breaks = c(seq(0, 85, by = 5), Inf), per_capita_household_size = fairfield_hh, school_demographics = conmat_original_school_demographics, work_demographics = conmat_original_work_demographics ) # or use non-symmetric model terms contact_model_pred_est <- estimate_setting_contacts( contact_data_list = get_polymod_setting_data(), survey_population = get_polymod_population(), prediction_population = abs_age_lga("Fairfield (C)"), age_breaks = c(seq(0, 85, by = 5), Inf), per_capita_household_size = fairfield_hh, symmetrical = FALSE ) ## End(Not run)
Uses estimate_setting_contacts() to fit a contact model on the data from
polymod and later extrapolate on to a desired population. Note that this
function is parallelisable with future, and will be impacted by any
future plans provided.
extrapolate_polymod( population, age_breaks = c(seq(0, 75, by = 5), Inf), per_capita_household_size = NULL )extrapolate_polymod( population, age_breaks = c(seq(0, 75, by = 5), Inf), per_capita_household_size = NULL )
population |
a |
age_breaks |
vector depicting age values. Default value is
|
per_capita_household_size |
Optional (defaults to NULL). When set, it
adjusts the household contact matrix by some per capita household size.
To set it, provide a single number, the per capita household size. More
information is provided below in Details. See
|
Also note that since this model uses the already fit polymod_setting_models
data, which has been fit using symmetrical model terms, if you want to
fit a model with asymmetric model terms, you will need to go through
the full process of building new models. You can find this detail in last
section of the vignette "example pipeline".
Returns setting-specific and combined contact matrices for the desired ages.
## Not run: polymod_population <- get_polymod_population() synthetic_settings_5y_polymod <- extrapolate_polymod( population = polymod_population ) synthetic_settings_5y_polymod synthetic_settings_5y_fairfield <- extrapolate_polymod( population = abs_age_lga("Fairfield (C)") ) synthetic_settings_5y_fairfield ## End(Not run)## Not run: polymod_population <- get_polymod_population() synthetic_settings_5y_polymod <- extrapolate_polymod( population = polymod_population ) synthetic_settings_5y_polymod synthetic_settings_5y_fairfield <- extrapolate_polymod( population = abs_age_lga("Fairfield (C)") ) synthetic_settings_5y_fairfield ## End(Not run)
A dataset containing data digitised from "The impact of
SARS-CoV-2 vaccination on Alpha & Delta variant transmission", by David W
Eyre, Donald Taylor, Mark Purver, David Chapman, Tom Fowler, Koen B Pouwels,
A Sarah Walker, Tim EA Peto
(doi:10.1101/2021.09.28.21264260. The figures were taken
from
https://www.medrxiv.org/content/10.1101/2021.09.28.21264260v1.full-text,
and the code to digitise these figures is in data-raw under
"read_eyre_transmission_probabilities.R". When using this data, ensure that
you cite the original authors at 'Eyre, D. W., Taylor, D., Purver, M., Chapman, D., Fowler, T., Pouwels, K. B., Walker, A. S., & Peto, T. E. (2021). The impact of SARS-CoV-2 vaccination on Alpha & Delta variant transmission (Preprint). Infectious Diseases (except HIV/AIDS). https://doi.org/10.1101/2021.09.28.21264260'
eyre_transmission_probabilitieseyre_transmission_probabilities
A data frame of the probability of transmission from a case to a contact. There are 40,804 rows and 6 variables.
"household", "household_visitor", "work_education", or "events_activities"
from 0 to 100
from ages 0 to 100
If case is between ages 0-4, in 5 year bins up to 100
If contact is between ages 0-4, in 5 year bins up to 100
probability of transmission. Value is 0 - 1
## Not run: # plot this library(ggplot2) library(stringr) eyre_transmission_probabilities %>% group_by( setting, case_age_5y, contact_age_5y ) %>% summarise( across( probability, mean ), .groups = "drop" ) %>% rename( case_age = case_age_5y, contact_age = contact_age_5y ) %>% mutate( across( ends_with("age"), ~ factor(.x, levels = str_sort( unique(.x), numeric = TRUE ) ) ) ) %>% ggplot( aes( x = case_age, y = contact_age, fill = probability ) ) + facet_wrap(~setting) + geom_tile() + scale_fill_viridis() + coord_fixed() + theme_minimal() + theme( axis.text = element_text(angle = 45, hjust = 1) ) ## End(Not run)## Not run: # plot this library(ggplot2) library(stringr) eyre_transmission_probabilities %>% group_by( setting, case_age_5y, contact_age_5y ) %>% summarise( across( probability, mean ), .groups = "drop" ) %>% rename( case_age = case_age_5y, contact_age = contact_age_5y ) %>% mutate( across( ends_with("age"), ~ factor(.x, levels = str_sort( unique(.x), numeric = TRUE ) ) ) ) %>% ggplot( aes( x = case_age, y = contact_age, fill = probability ) ) + facet_wrap(~setting) + geom_tile() + scale_fill_viridis() + coord_fixed() + theme_minimal() + theme( axis.text = element_text(angle = 45, hjust = 1) ) ## End(Not run)
fits a gam model for each setting on the survey population data & the
setting wise contact data. The underlying method is described in more
detail in fit_single_contact_model(). The models can be fit in parallel,
see the examples. Note that this function is parallelisable with future,
and will be impacted by any future plans provided.
fit_setting_contacts( contact_data_list, population, symmetrical = TRUE, school_demographics = NULL, work_demographics = NULL )fit_setting_contacts( contact_data_list, population, symmetrical = TRUE, school_demographics = NULL, work_demographics = NULL )
contact_data_list |
A list of dataframes, each containing information
on the setting (home, work, school, other), age_from, age_to,
the number of contacts, and the number of participants. Example data
can be retrieved with |
population |
|
symmetrical |
whether to enforce symmetrical terms in the model.
Defaults to TRUE. See |
school_demographics |
(optional) defaults to census average proportion
at school. You can provide a dataset with columns, "age" (numeric), and
"school_fraction" (0-1), if you would like to specify these
details. See |
work_demographics |
(optional) defaults to census average proportion
employed. You can provide a dataset with columns, "age" (numeric), and
"work_fraction", if you would like to specify these details. See
|
list of fitted gam models - one for each setting provided
Nicholas Tierney
# These aren't being run as they take too long to fit ## Not run: contact_model <- fit_setting_contacts( contact_data_list = get_polymod_setting_data(), population = get_polymod_population() ) # can fit the model in parallel library(future) plan(multisession, workers = 4) polymod_setting_data <- get_polymod_setting_data() polymod_population <- get_polymod_population() contact_model <- fit_setting_contacts( contact_data_list = polymod_setting_data, population = polymod_population ) # you can specify your own population data for school and work demographics contact_model_diff_data <- fit_setting_contacts( contact_data_list = polymod_setting_data, population = polymod_population, school_demographics = conmat_original_school_demographics, work_demographics = conmat_original_work_demographics ) ## End(Not run)# These aren't being run as they take too long to fit ## Not run: contact_model <- fit_setting_contacts( contact_data_list = get_polymod_setting_data(), population = get_polymod_population() ) # can fit the model in parallel library(future) plan(multisession, workers = 4) polymod_setting_data <- get_polymod_setting_data() polymod_population <- get_polymod_population() contact_model <- fit_setting_contacts( contact_data_list = polymod_setting_data, population = polymod_population ) # you can specify your own population data for school and work demographics contact_model_diff_data <- fit_setting_contacts( contact_data_list = polymod_setting_data, population = polymod_population, school_demographics = conmat_original_school_demographics, work_demographics = conmat_original_work_demographics ) ## End(Not run)
This is the workhorse of the conmat package, and is typically
used inside fit_setting_contacts(). It predicts the contact rate between
all age bands (the contact rate between ages 0 and 1, 0 and 2, 0 and 3,
and so on), for a specified setting, with specific terms being added for
given settings. See "details" for further information.
fit_single_contact_model( contact_data, population, symmetrical = TRUE, school_demographics = NULL, work_demographics = NULL )fit_single_contact_model( contact_data, population, symmetrical = TRUE, school_demographics = NULL, work_demographics = NULL )
contact_data |
dataset with columns |
population |
|
symmetrical |
whether to enforce symmetrical terms in the model.
Defaults to TRUE. See |
school_demographics |
(optional) defaults to census average proportion
at school. You can provide a dataset with columns, "age" (numeric), and
"school_fraction" (0-1), if you would like to specify these
details. See |
work_demographics |
(optional) defaults to census average proportion
employed. You can provide a dataset with columns, "age" (numeric), and
"work_fraction", if you would like to specify these details. See
|
The model fit is a Generalised Additive Model (GAM). We provide two "modes" for model fitting. Either using "symmetric" or "non-symmetric" model predictor terms with the logical variance "symmetrical", which is set to TRUE by default. We recommend using the "symmetrical" terms as it reflects the fact that contacts are symmetric - person A having contact with person B means person B has had contact with person A. We've included a variety of terms to account for assortativity with age, where people of similar ages have more contact with each other. And included terms to account for intergenerational contact patterns, where parents and grandparents will interact with their children and grand children. These terms are fit with a smoothing function. Specifically, the relevant code looks like this:
# abs(age_from - age_to) s(gam_age_offdiag) + # abs(age_from - age_to)^2 s(gam_age_offdiag_2) + # abs(age_from * age_to) s(gam_age_diag_prod) + # abs(age_from + age_to) s(gam_age_diag_sum) + # pmax(age_from, age_to) s(gam_age_pmax) + # pmin(age_from, age_to) s(gam_age_pmin)
We also include predictors for the probability of attending school, and
attending work. These are computed as the probability that a person goes
to the same school/work, proportional to the increase in contacts due to
attendance. These terms are calculated from estimated proportion of
people in age groups attending school and work. See
add_modelling_features() for more details.
Finally, we include two offset terms so that we estimate the contact rate,
that is the contacts per capita, instead of the number of contacts. These
offset terms are log(contactable_population), and
log(contactable_population_school) when the model is fit to a school
setting. The contactable population is estimated as the interpolated
1 year ages from the data. For schools this is the contactable population
weighted by the proportion of the population attending school.
This leaves us with a model that looks like so:
mgcv::bam(
formula = contacts ~
# abs(age_from - age_to)
s(gam_age_offdiag) +
# abs(age_from - age_to)^2
s(gam_age_offdiag_2) +
# abs(age_from * age_to)
s(gam_age_diag_prod) +
# abs(age_from + age_to)
s(gam_age_diag_sum) +
# pmax(age_from, age_to)
s(gam_age_pmax) +
# pmin(age_from, age_to)
s(gam_age_pmin) +
school_probability +
work_probability +
offset(log_contactable_population) +
# or for school settings
# offset(log_contactable_population_school)
family = stats::poisson,
offset = log(participants),
data = population_data
)
But if the term symmetrical = FALSE is used, you get:
mgcv::bam(
formula = contacts ~
s(age_to) +
s(age_from) +
s(abs(age_from - age_to)) +
s(abs(age_from - age_to), age_from) +
school_probability +
work_probability +
offset(log_contactable_population) +
# or for school settings
# offset(log_contactable_population_school)
family = stats::poisson,
offset = log(participants),
data = population_data
)
single model
example_contact <- get_polymod_contact_data(setting = "home") example_contact example_population <- get_polymod_population() library(dplyr) example_contact_20 <- example_contact %>% filter( age_to <= 20, age_from <= 20 ) my_mod <- fit_single_contact_model( contact_data = example_contact_20, population = example_population ) # you can specify your own population data for school and work demographics my_mod_diff_data <- fit_single_contact_model( contact_data = example_contact_20, population = example_population, school_demographics = conmat_original_school_demographics, work_demographics = conmat_original_work_demographics )example_contact <- get_polymod_contact_data(setting = "home") example_contact example_population <- get_polymod_population() library(dplyr) example_contact_20 <- example_contact %>% filter( age_to <= 20, age_from <= 20 ) my_mod <- fit_single_contact_model( contact_data = example_contact_20, population = example_population ) # you can specify your own population data for school and work demographics my_mod_diff_data <- fit_single_contact_model( contact_data = example_contact_20, population = example_population, school_demographics = conmat_original_school_demographics, work_demographics = conmat_original_work_demographics )
Once infected, a person can transmit an infectious disease to
another, creating generations of infected individuals. We can define a
matrix describing the number of newly infected individuals in given
categories, such as age, for consecutive generations. This matrix is
called a "next generation matrix" (NGM). We can generate an NGM from two
sources - a conmat_population object (such as the output from
abs_age_lga()), or a conmat_setting_prediction_matrix, which is the
output from extrapolate_polymod() or predict_setting_contacts().
generate_ngm(x, age_breaks, R_target, setting_transmission_matrix, ...) ## S3 method for class 'conmat_setting_prediction_matrix' generate_ngm( x, age_breaks, R_target, setting_transmission_matrix = NULL, per_capita_household_size = NULL, ..., lga_name, state_name ) ## S3 method for class 'conmat_population' generate_ngm( x, age_breaks, R_target, setting_transmission_matrix = NULL, per_capita_household_size = NULL, ..., lga_name, state_name )generate_ngm(x, age_breaks, R_target, setting_transmission_matrix, ...) ## S3 method for class 'conmat_setting_prediction_matrix' generate_ngm( x, age_breaks, R_target, setting_transmission_matrix = NULL, per_capita_household_size = NULL, ..., lga_name, state_name ) ## S3 method for class 'conmat_population' generate_ngm( x, age_breaks, R_target, setting_transmission_matrix = NULL, per_capita_household_size = NULL, ..., lga_name, state_name )
x |
data input - could be a |
age_breaks |
vector depicting age values with the highest age depicted
as |
R_target |
target reproduction number |
setting_transmission_matrix |
default is NULL, which calculates the transmission
matrix using |
... |
extra arguments, currently not used |
per_capita_household_size |
default is NULL - which defaults to |
lga_name |
now defunct, but capturing arguments for informative error |
state_name |
now defunct, but capturing arguments for informative error |
The NGM can be used to calculate the expected number of secondary infections in a given age group. Given certain age breaks, we compute the unscaled next generation matrices for that location across different settings & age groups using the contact rates extrapolated from POLYMOD survey data on the specified location, adjusted by the per capita household size and the setting-specific relative per-contact transmission probability matrices for the same age groups. These NGMs are then scaled according to a target reproduction number (which is provided as an argument) using the ratio of the desired R0 and the R0 of the NGM for the combination of all settings. The R0 of the combination of all settings is obtained by calculating the unique, positive eigen value of the combination NGM. This ratio is then used to scale all the setting specific NGMs.
When using a setting prediction contact matrix (such as one generated
by extrapolate_polymod, with class conmat_setting_prediction_matrix),
the age breaks specified in generate_ngm must be the same as the age
breaks specified in the synthetic contact matrix, otherwise it will error
as it is trying to multiple incompatible matrices.
## Not run: perth <- abs_age_lga("Perth (C)") perth_hh <- get_abs_per_capita_household_size(lga = "Perth (C)") age_breaks_0_80_plus <- c(seq(0, 80, by = 5), Inf) # you can also run this without `per_capita_household_size` perth_ngm_lga <- generate_ngm( perth, age_breaks = age_breaks_0_80_plus, per_capita_household_size = perth_hh, R_target = 1.5 ) perth_contact <- extrapolate_polymod( perth, per_capita_household_size = perth_hh ) perth_ngm <- generate_ngm( perth_contact, age_breaks = age_breaks_0_80_plus, R_target = 1.5 ) # using our own transmission matrix new_transmission_matrix <- get_setting_transmission_matrices( age_breaks = age_breaks_0_80_plus, # is normally 0.5 asymptomatic_relative_infectiousness = 0.75 ) new_transmission_matrix perth_ngm_0_80_new_tmat <- generate_ngm( perth_contact, age_breaks = age_breaks_0_80_plus, R_target = 1.5, setting_transmission_matrix = new_transmission_matrix ) ## End(Not run) # examples not run as they take a long time ## Not run: perth <- abs_age_lga("Perth (C)") perth_contact <- extrapolate_polymod(perth) generate_ngm(perth_contact, age_breaks = c(seq(0, 85, by = 5), Inf)) ## End(Not run)## Not run: perth <- abs_age_lga("Perth (C)") perth_hh <- get_abs_per_capita_household_size(lga = "Perth (C)") age_breaks_0_80_plus <- c(seq(0, 80, by = 5), Inf) # you can also run this without `per_capita_household_size` perth_ngm_lga <- generate_ngm( perth, age_breaks = age_breaks_0_80_plus, per_capita_household_size = perth_hh, R_target = 1.5 ) perth_contact <- extrapolate_polymod( perth, per_capita_household_size = perth_hh ) perth_ngm <- generate_ngm( perth_contact, age_breaks = age_breaks_0_80_plus, R_target = 1.5 ) # using our own transmission matrix new_transmission_matrix <- get_setting_transmission_matrices( age_breaks = age_breaks_0_80_plus, # is normally 0.5 asymptomatic_relative_infectiousness = 0.75 ) new_transmission_matrix perth_ngm_0_80_new_tmat <- generate_ngm( perth_contact, age_breaks = age_breaks_0_80_plus, R_target = 1.5, setting_transmission_matrix = new_transmission_matrix ) ## End(Not run) # examples not run as they take a long time ## Not run: perth <- abs_age_lga("Perth (C)") perth_contact <- extrapolate_polymod(perth) generate_ngm(perth_contact, age_breaks = c(seq(0, 85, by = 5), Inf)) ## End(Not run)
This function calculates a next generation matrix (NGM)
based on state or LGA data from the Australian Bureau of Statistics (ABS).
For full details see generate_ngm().
generate_ngm_oz( state_name = NULL, lga_name = NULL, age_breaks, R_target, setting_transmission_matrix = NULL )generate_ngm_oz( state_name = NULL, lga_name = NULL, age_breaks, R_target, setting_transmission_matrix = NULL )
state_name |
target Australian state name in abbreviated form, such as "QLD", "NSW", or "TAS" |
lga_name |
target Australian local government area (LGA) name, such
as "Fairfield (C)". See |
age_breaks |
vector depicting age values with the highest age depicted
as |
R_target |
target reproduction number |
setting_transmission_matrix |
default is NULL, which calculates the transmission
matrix using |
# don't run as both together takes a long time to run ## Not run: ngm_nsw <- generate_ngm_oz( state_name = "NSW", age_breaks = c(seq(0, 85, by = 5), Inf), R_target = 1.5 ) ngm_fairfield <- generate_ngm_oz( lga_name = "Fairfield (C)", age_breaks = c(seq(0, 85, by = 5), Inf), R_target = 1.5 ) ## End(Not run)# don't run as both together takes a long time to run ## Not run: ngm_nsw <- generate_ngm_oz( state_name = "NSW", age_breaks = c(seq(0, 85, by = 5), Inf), R_target = 1.5 ) ngm_fairfield <- generate_ngm_oz( lga_name = "Fairfield (C)", age_breaks = c(seq(0, 85, by = 5), Inf), R_target = 1.5 ) ## End(Not run)
Get household size distribution based on state or LGA name
get_abs_household_size_distribution(state = NULL, lga = NULL)get_abs_household_size_distribution(state = NULL, lga = NULL)
state |
target Australian state name in abbreviated form, such as "QLD", "NSW", or "TAS" |
lga |
target Australian local government area (LGA) name, such as "Fairfield (C)". See
|
returns a data frame with household size distributions of a specific state or LGA
get_abs_household_size_distribution(lga = "Fairfield (C)") get_abs_household_size_distribution(state = "NSW") ## Not run: # cannot specify both state and LGA get_abs_household_size_distribution(state = "NSW", lga = "Fairfield (C)") ## End(Not run)get_abs_household_size_distribution(lga = "Fairfield (C)") get_abs_household_size_distribution(state = "NSW") ## Not run: # cannot specify both state and LGA get_abs_household_size_distribution(state = "NSW", lga = "Fairfield (C)") ## End(Not run)
Get population associated with each household size in an LGA or a state
get_abs_household_size_population(state = NULL, lga = NULL)get_abs_household_size_population(state = NULL, lga = NULL)
state |
target Australian state name in abbreviated form, such as "QLD", "NSW", or "TAS" |
lga |
target Australian local government area (LGA) name, such as "Fairfield (C)". See
|
returns a data frame with household size and the population associated with it in each LGA or state.
get_abs_household_size_population(state = "NSW")get_abs_household_size_population(state = "NSW")
Get per capita household size based on state or LGA name
get_abs_per_capita_household_size(state = NULL, lga = NULL)get_abs_per_capita_household_size(state = NULL, lga = NULL)
state |
state name |
lga |
lga name |
Numeric of length 1 - the per capita household size for a given state or LGA.
Nick Golding
get_abs_per_capita_household_size(lga = "Fairfield (C)") get_abs_per_capita_household_size(state = "NSW") ## Not run: # cannot specify both state and LGA get_abs_per_capita_household_size(state = "NSW", lga = "Fairfield (C)") ## End(Not run)get_abs_per_capita_household_size(lga = "Fairfield (C)") get_abs_per_capita_household_size(state = "NSW") ## Not run: # cannot specify both state and LGA get_abs_per_capita_household_size(state = "NSW", lga = "Fairfield (C)") ## End(Not run)
Get household size distribution based on LGA name
get_abs_per_capita_household_size_lga(lga = NULL)get_abs_per_capita_household_size_lga(lga = NULL)
lga |
target Australian local government area (LGA) name, such as "Fairfield (C)". See
|
returns a numeric value depicting the per capita household size of the specified LGA
get_abs_per_capita_household_size_lga(lga = "Fairfield (C)")get_abs_per_capita_household_size_lga(lga = "Fairfield (C)")
Get household size distribution based on state name
get_abs_per_capita_household_size_state(state = NULL)get_abs_per_capita_household_size_state(state = NULL)
state |
target Australian state name in abbreviated form, such as "QLD", "NSW", or "TAS" |
returns a numeric value depicting the per capita household size of the specified state
get_abs_per_capita_household_size_state(state = "NSW")get_abs_per_capita_household_size_state(state = "NSW")
This function returns an interpolating function to get
populations in 1y age increments from chunkier distributions produced by
socialmixr::wpp_age().
get_age_population_function(data, ...) ## S3 method for class 'conmat_population' get_age_population_function(data = population, ...) ## S3 method for class 'data.frame' get_age_population_function( data = population, age_col = lower.age.limit, pop_col = population, ... )get_age_population_function(data, ...) ## S3 method for class 'conmat_population' get_age_population_function(data = population, ...) ## S3 method for class 'data.frame' get_age_population_function( data = population, age_col = lower.age.limit, pop_col = population, ... )
data |
dataset containing information on population of a given age/age group |
... |
extra arguments |
age_col |
bare variable name for the column with age information |
pop_col |
bare variable name for the column with population information |
The function first prepares the data to fit a smoothing spline to the data for ages below the maximum age. It arranges the data by the lower limit of the age group to obtain the bin width/differences of the lower age limits. The mid point of the bin width is later added to the ages and the population is scaled as per the bin widths. The maximum age is later obtained and the populations for different above and below are filtered out along with the sum of populations with and without maximum age. A cubic smoothing spline is then fitted to the data for ages below the maximum with predictor variable as the ages with the mid point of the bins added to it where as the response variable is the log-scaled population. Using the smoothing spline fit, the predicted population of ages 0 to 200 is obtained and the predicted population is adjusted further using a ratio of the sum of the population across all ages from the data and predicted population. The ratio is based on whether the ages are under the maximum age as the total population across all ages differs for ages above and below the maximum age. The maximum age population is adjusted further to drop off smoothly, based on the weights. The final population is then linearly extrapolated over years past the upper bound from the data. For ages above the maximum age from data, the population is calculated as a weighted population of the maximum age that depends on the years past the upper bound. Older ages would have lower weights, therefore lower population.
An interpolating function to get populations in 1y age increments
polymod_pop <- get_polymod_population() polymod_pop # But these ages and populations are binned every 5 years. So we can now # provide a specified age and get the estimated population for that 1 year # age group. First we create the new function like so age_pop_function <- get_age_population_function( data = polymod_pop ) # Then we pass it a year to get the estimated population for a particular age age_pop_function(4) # Or a vector of years, to get the estimated population for a particular age # range age_pop_function(1:4) # Notice that we get a _pretty similar_ number of 0-4 if we sum it up, as # the first row of the table: head(polymod_pop, 1) sum(age_pop_function(age = 0:4)) # Usage in dplyr library(dplyr) example_df <- slice_head(abs_education_state, n = 5) example_df %>% mutate(population_est = age_pop_function(age))polymod_pop <- get_polymod_population() polymod_pop # But these ages and populations are binned every 5 years. So we can now # provide a specified age and get the estimated population for that 1 year # age group. First we create the new function like so age_pop_function <- get_age_population_function( data = polymod_pop ) # Then we pass it a year to get the estimated population for a particular age age_pop_function(4) # Or a vector of years, to get the estimated population for a particular age # range age_pop_function(1:4) # Notice that we get a _pretty similar_ number of 0-4 if we sum it up, as # the first row of the table: head(polymod_pop, 1) sum(age_pop_function(age = 0:4)) # Usage in dplyr library(dplyr) example_df <- slice_head(abs_education_state, n = 5) example_df %>% mutate(population_est = age_pop_function(age))
Provides contact and participant POLYMOD data from selected countries. It impute missing contact ages via one of three methods:
imputing contact ages from a random uniform distribution from the range of ages. 2) using the average of the ages, 3) removal of those participants. The contact settings are then classified as "home", "school", "work" and "others", where "others" include locations such as leisure, transport or other places. The participants with missing contact ages or settings are removed, and the number of contacts per participant and contact age from ages 0-100 are obtained for various countries and settings.
get_polymod_contact_data( setting = c("all", "home", "work", "school", "other"), countries = c("Belgium", "Finland", "Germany", "Italy", "Luxembourg", "Netherlands", "Poland", "United Kingdom"), ages = 0:100, contact_age_imputation = c("sample", "mean", "remove_participant") )get_polymod_contact_data( setting = c("all", "home", "work", "school", "other"), countries = c("Belgium", "Finland", "Germany", "Italy", "Luxembourg", "Netherlands", "Poland", "United Kingdom"), ages = 0:100, contact_age_imputation = c("sample", "mean", "remove_participant") )
setting |
Which setting to extract data from. Default is all settings. Options are: "all", "home", "work", "school", and "other". |
countries |
countries to extract data from. Default is all countries from this list: "Belgium", "Finland", "Germany", "Italy", "Luxembourg", "Netherlands", "Poland", and "United Kingdom". |
ages |
Which ages to return. Default is ages 0 to 100. |
contact_age_imputation |
How to handle age when it is missing. Choose one of three methods: 1) "sample", which imputes contact ages from a random uniform distribution from the range of ages. 2) "mean", use the average of the ages, 3) "remove_participant" removal of those participants. Default is "sample". |
A data.frame with columns: "setting" (all, work, home, etc. as specified in "setting" argument); "age_from" - the age of the participant; "age_to" - the age of the person the participant had contact with; "contacts" the number of contacts that person had; "participants" the number of participants in that row.
get_polymod_contact_data() get_polymod_contact_data(setting = "home") get_polymod_contact_data(countries = "Belgium") get_polymod_contact_data(countries = c("Belgium", "Italy")) get_polymod_contact_data(ages = 0:50) get_polymod_contact_data(contact_age_imputation = "sample") get_polymod_contact_data(contact_age_imputation = "mean") get_polymod_contact_data(contact_age_imputation = "remove_participant")get_polymod_contact_data() get_polymod_contact_data(setting = "home") get_polymod_contact_data(countries = "Belgium") get_polymod_contact_data(countries = c("Belgium", "Italy")) get_polymod_contact_data(ages = 0:50) get_polymod_contact_data(contact_age_imputation = "sample") get_polymod_contact_data(contact_age_imputation = "mean") get_polymod_contact_data(contact_age_imputation = "remove_participant")
Convenience function to help get the per capita household size.
This is calculated as mean(socialmixr::polymod$participants$hh_size).
get_polymod_per_capita_household_size()get_polymod_per_capita_household_size()
number, 3.248971
Nicholas Tierney
returns the polymod-average population age distribution in
5y increments (weight country population distributions by number of
participants). Note that we don't want to weight by survey age
distributions for this, since the total number of participants
represents the sampling. It uses the participant data from the polymod
survey as well as the age specific population data from socialmixr
R package to return the age specific average population of different,
countries weighted by the number of participants from those countries who
participated in the polymod survey.
get_polymod_population( countries = c("Belgium", "Finland", "Germany", "Italy", "Luxembourg", "Netherlands", "Poland", "United Kingdom") )get_polymod_population( countries = c("Belgium", "Finland", "Germany", "Italy", "Luxembourg", "Netherlands", "Poland", "United Kingdom") )
countries |
countries to extract data from. Default is to get: Belgium, Finland, Germany, Italy, Luxembourg, Netherlands, Poland, and United Kingdom. |
A conmat_population data frame with two columns: lower.age.limit
and population
get_polymod_population() get_polymod_population("Belgium") get_polymod_population("United Kingdom") get_polymod_population("Italy")get_polymod_population() get_polymod_population("Belgium") get_polymod_population("United Kingdom") get_polymod_population("Italy")
get_polymod_setting_data() acts as an extension of
get_polymod_contact_data(), and extracts the setting wise contact data
on the desired country, as a list.
get_polymod_setting_data( countries = c("Belgium", "Finland", "Germany", "Italy", "Luxembourg", "Netherlands", "Poland", "United Kingdom") )get_polymod_setting_data( countries = c("Belgium", "Finland", "Germany", "Italy", "Luxembourg", "Netherlands", "Poland", "United Kingdom") )
countries |
countries to extract data from |
A list of data frames, of the polymod data. One list per setting: "home", "work", "school", and "other".
get_polymod_setting_data() get_polymod_setting_data("Belgium")get_polymod_setting_data() get_polymod_setting_data("Belgium")
Given some age breaks, return a named list of matrices containing age-specific relative per-contact transmission probability matrices for each of 4 settings: home, school, work, other. These can be combined with contact matrices to produce setting-specific relative next generation matrices (NGMs). These can be scaled to match a required reproduction number based on the dominant eigenvalue of the all-settings NGM (the elementwise sum of all setting NGMs).
get_setting_transmission_matrices( age_breaks = c(seq(0, 80, by = 5), Inf), asymptomatic_relative_infectiousness = 0.5, susceptibility_estimate = c("davies_updated", "davies_original") )get_setting_transmission_matrices( age_breaks = c(seq(0, 80, by = 5), Inf), asymptomatic_relative_infectiousness = 0.5, susceptibility_estimate = c("davies_updated", "davies_original") )
age_breaks |
vector of age breaks, defaults to
|
asymptomatic_relative_infectiousness |
the assumed ratio of onward infectiousness between asymptomatic and symptomatic cases. This represents the infectiousness of asymptomatic relative to symptomatic. Default value is 0.5, which means the asymptomatic cases are 50% less infectious than symptomatic cases. |
susceptibility_estimate |
Which estimate to use for susceptibility by age. Either, the smoothed original Davies et al estimates, "davies_original" or, the set updated to match UK under-16 infections (the default), "davies_updated". |
These matrices are created from: an estimate of the clinical fraction for each age (inferred by applying a smoothing spline to the mean estimates from Davies et al.); an assumption of the infectiousness of asymptomatics relative to symptomatics (provided as an argument); estimates of the relative susceptibility to infection of individuals of different ages, inferred from a smoothing-spline estimate of the mean relative susceptibility estimate from Davies et al., combined with a re-estimation of the susceptibility profile for under-16s, estimated in a similar way but to the age-distribution of infections in England from the UK ONS prevalence survey (rather than case counts with may undercount children), assuming the above clinical fraction estimates, and accounting for vaccination, reduced mixing, and reduced transmissibility in work and other settings due to hygiene behaviour; and estimates of the relative transmissibility in household vs non-household settings - scaled linearly for non-household transmission and binomially for household transmission (so that onward infections do not to exceed the number of other household members).
When using this data, ensure that you cite this package, and the original authors of the paper from which these estimates were derived:
Davies, N.G., Klepac, P., Liu, Y. et al. Age-dependent effects in the transmission and control of COVID-19 epidemics. Nat Med 26, 1205–1211 (2020). https://doi.org/10.1038/s41591-020-0962-9
list of matrices, containing the relative per-contact transmission probability for each setting
## Not run: # fit polymod model setting_models <- fit_setting_contacts( contact_data_list = get_polymod_setting_data(), population = get_polymod_population() ) # define age breaks for prediction age_breaks <- c(seq(0, 80, by = 5), Inf) # define a new population age distribution to predict to fairfield <- abs_age_lga("Fairfield (C)") # predict setting-specific contact matrices to a new population contact_matrices <- predict_setting_contacts( population = fairfield, contact_model = setting_models, age_breaks = age_breaks ) # remove the 'all' matrix, keep the other four settings contact_matrices <- contact_matrices[c("home", "school", "work", "other")] # get setting-specific per-contact transmission rate matrices for the same # age aggregations transmission_matrices <- get_setting_transmission_matrices( age_breaks = age_breaks ) # combine them to get setting-specific (unscaled) next-generation matrices next_generation_matrices <- mapply( FUN = `*`, contact_matrices, transmission_matrices, SIMPLIFY = FALSE ) # get the all-settings NGM ngm_overall <- Reduce("+", next_generation_matrices) ## End(Not run)## Not run: # fit polymod model setting_models <- fit_setting_contacts( contact_data_list = get_polymod_setting_data(), population = get_polymod_population() ) # define age breaks for prediction age_breaks <- c(seq(0, 80, by = 5), Inf) # define a new population age distribution to predict to fairfield <- abs_age_lga("Fairfield (C)") # predict setting-specific contact matrices to a new population contact_matrices <- predict_setting_contacts( population = fairfield, contact_model = setting_models, age_breaks = age_breaks ) # remove the 'all' matrix, keep the other four settings contact_matrices <- contact_matrices[c("home", "school", "work", "other")] # get setting-specific per-contact transmission rate matrices for the same # age aggregations transmission_matrices <- get_setting_transmission_matrices( age_breaks = age_breaks ) # combine them to get setting-specific (unscaled) next-generation matrices next_generation_matrices <- mapply( FUN = `*`, contact_matrices, transmission_matrices, SIMPLIFY = FALSE ) # get the all-settings NGM ngm_overall <- Reduce("+", next_generation_matrices) ## End(Not run)
This function is the opposite of predictions_to_matrix(). It
converts a wide matrix into a long data frame. It is mostly used within
plotting functions.
matrix_to_predictions(contact_matrix)matrix_to_predictions(contact_matrix)
contact_matrix |
square matrix with age group to and from information in the row and column names. |
data.frame with columns age_group_to, age_group_from, and
contacts.
fairfield <- abs_age_lga("Fairfield (C)") # We can convert the predictions into a matrix fairfield_school_contacts <- predict_contacts( model = polymod_setting_models$school, population = fairfield, age_breaks = c(0, 5, 10, 15, Inf) ) fairfield_school_contacts fairfield_school_mat <- predictions_to_matrix(fairfield_school_contacts) fairfield_school_mat matrix_to_predictions(fairfield_school_mat)fairfield <- abs_age_lga("Fairfield (C)") # We can convert the predictions into a matrix fairfield_school_contacts <- predict_contacts( model = polymod_setting_models$school, population = fairfield, age_breaks = c(0, 5, 10, 15, Inf) ) fairfield_school_contacts fairfield_school_mat <- predictions_to_matrix(fairfield_school_contacts) fairfield_school_mat matrix_to_predictions(fairfield_school_mat)
A matrix that knows about its age breaks - which are by default provided as its rownames. Mostly intended for internal use.
new_age_matrix(matrix, age_breaks)new_age_matrix(matrix, age_breaks)
matrix |
numeric matrix |
age_breaks |
character vector of age breaks, by default the rownames. |
matrix with age breaks attribute
age_break_names <- c("[0,5)", "[5,10)", "[10, 15)") age_mat <- matrix( runif(9), nrow = 3, ncol = 3, dimnames = list( age_break_names, age_break_names ) ) new_age_matrix( age_mat, age_breaks = age_break_names )age_break_names <- c("[0,5)", "[5,10)", "[10, 15)") age_mat <- matrix( runif(9), nrow = 3, ncol = 3, dimnames = list( age_break_names, age_break_names ) ) new_age_matrix( age_mat, age_breaks = age_break_names )
Establish new BGM setting data
new_ngm_setting_matrix(list_matrix, raw_eigenvalue, scaling, age_breaks)new_ngm_setting_matrix(list_matrix, raw_eigenvalue, scaling, age_breaks)
list_matrix |
list of matrices |
raw_eigenvalue |
the raw eigenvalue |
scaling |
scaling factor |
age_breaks |
vector of age breaks |
object with additional (primary) class "ngm_setting_matrix", and attributes for "age_breaks", "scaling", and "raw_eigenvalue".
Establish new setting data
new_setting_data(list_df)new_setting_data(list_df)
list_df |
list of data frames |
object with additional (primary) class "setting data" and an "age_breaks attribute.
Returns the per capita household size for a location given
its household size distribution. See get_abs_household_size_distribution()
function for retrieving household size distributions for a given place.
per_capita_household_size( household_data, household_size_col = household_size, n_people_col = n_people )per_capita_household_size( household_data, household_size_col = household_size, n_people_col = n_people )
household_data |
data set with information on the household size distribution of specific state or LGA. |
household_size_col |
bare variable name of the column depicting the
household size. Default is 'household_size' from
|
n_people_col |
bare variable name of the column depicting the total
number of people belonging to the respective household size. Default is
'n_people' from |
Numeric of length 1 - the per capita household size for a given state or LGA.
Nick Golding
demo_data <- get_abs_household_size_population(lga = "Fairfield (C)") demo_data per_capita_household_size( household_data = demo_data, household_size_col = household_size, n_people_col = n_people )demo_data <- get_abs_household_size_population(lga = "Fairfield (C)") demo_data per_capita_household_size( household_data = demo_data, household_size_col = household_size, n_people_col = n_people )
socialmixr)A dataset containing social mixing diary data from 8 European countries: Belgium, Germany, Finland, Great Britain, Italy, Luxembourg, The Netherlands and Poland.
polymodpolymod
A list of two data frames:
the study participant, with age, country, year and day of the week (starting with 1 = Monday)
reported contacts of the study participants. The variable phys_contact has two levels (1 denotes physical contact while 2 denotes non-physical contact), duration_multi has five levels (1 is less than 5 minutes while 5 is more than 4 hours, increasing in the order found in Figure 1 in Mossong et al.), and frequency_multi has five levels (1 is daily, 2 is weekly, 3 is monthly, 4 is less often, and 5 is first time)
All other variables are described on the Zenodo repository of the data, available at doi:10.5281/zenodo.1043437
This data has been sourced from the socialmixr package.
The Data are fully described in Mossong J, Hens N, Jit M, Beutels P, Auranen K, Mikolajczyk R, et al. (2008) Social Contacts and Mixing Patterns Relevant to the Spread of Infectious Diseases. PLoS Med 5(3): e74.
doi:10.1371/journal.pmed.0050074
A data object containing a list of fitted gam models
predicting the number of contacts in each of the four settings which are
"home","work","school" and "other". For more details on model fitting,
see fit_setting_contacts(). This object has been provided as data to
avoid recomputing a relatively common type of model for use with conmat.
polymod_setting_modelspolymod_setting_models
An object of class list of length 4.
## Not run: # code used to produce this data library(conmat) set.seed(2022 - 08 - 26) polymod_contact_data <- get_polymod_setting_data() polymod_survey_data <- get_polymod_population() polymod_setting_models <- fit_setting_contacts( contact_data_list = polymod_contact_data, # population = polymod_survey_data ) ## End(Not run)## Not run: # code used to produce this data library(conmat) set.seed(2022 - 08 - 26) polymod_contact_data <- get_polymod_setting_data() polymod_survey_data <- get_polymod_population() polymod_setting_models <- fit_setting_contacts( contact_data_list = polymod_contact_data, # population = polymod_survey_data ) ## End(Not run)
Predicts the expected contact rate over specified age breaks,
given some model of contact rate and population age structure.
This function is used internally in predict_setting_contacts(), which
performs this prediction across all settings (home, work, school, other),
and optionally performs an adjustment for per capita household size. You
can use predict_contacts() by itself, just be aware you will need to
separately apply a per capita household size adjustment if required. See
details below on adjust_household_contact_matrix for more information.
predict_contacts(model, population, age_breaks = c(seq(0, 75, by = 5), Inf))predict_contacts(model, population, age_breaks = c(seq(0, 75, by = 5), Inf))
model |
A single fitted model of contact rate (e.g.,
|
population |
a dataframe of age population information, with columns
indicating some lower age limit, and population, (e.g., |
age_breaks |
the ages to predict to. By default, the age breaks are 0-75 in 5 year groups. |
The population data is used to determine age range to predict
contact rates, and removes ages with zero population, so we do not
make predictions for ages with zero populations. Contact rates are
predicted yearly between the age groups, using predict_contacts_1y(),
then aggregates these predicted contacts using
aggregate_predicted_contacts(), which aggregates the predictions back to
the same resolution as the data, appropriately weighting the contact rate
by the population.
Regarding the adjust_household_contact_matrix function, we use
Per-capita household size instead of mean household size.
Per-capita household size is different to mean household size, as the
household size averaged over people in the population, not over
households, so larger households get upweighted. It is calculated by
taking a distribution of the number of households of each size in a
population, multiplying the size by the household by the household count
to get the number of people with that size of household, and computing
the population-weighted average of household sizes. We use per-capita
household size as it is a more accurate reflection of the average
number of household members a person in the population can have contact
with.
A dataframe with three columns: age_group_from, age_group_to,
and contacts. The age groups are factors, broken up into 5 year bins
[0,5), [5,10). The contact column is the predicted number of
contacts from the specified age group to the other one.
# If we have a model of contact rate at home, and age population structure # for an LGA, say, Fairfield, in NSW: polymod_setting_models$home fairfield <- abs_age_lga("Fairfield (C)") fairfield # We can predict the contact rate for Fairfield from the existing contact # data, say, between the age groups of 0-15 in 5 year bins for school: fairfield_school_contacts <- predict_contacts( model = polymod_setting_models$school, population = fairfield, age_breaks = c(0, 5, 10, 15, Inf) ) fairfield_school_contacts# If we have a model of contact rate at home, and age population structure # for an LGA, say, Fairfield, in NSW: polymod_setting_models$home fairfield <- abs_age_lga("Fairfield (C)") fairfield # We can predict the contact rate for Fairfield from the existing contact # data, say, between the age groups of 0-15 in 5 year bins for school: fairfield_school_contacts <- predict_contacts( model = polymod_setting_models$school, population = fairfield, age_breaks = c(0, 5, 10, 15, Inf) ) fairfield_school_contacts
Provides a predicted rate of contacts for contact ages.
Take an already fitted model of contact rate and predict the
estimated contact rate, and standard error, for all combinations of the
provided ages in 1 year increments. So if the minimum age is 5, and the
maximum age is 10, it will provide the estimated contact rate for all age
combinations: 5 and 5, 5 and 6 ... 5 and 10, and so on. This function is
used internally within predict_contacts(), and thus
predict_setting_contacts() as well, although it can be used by itself.
See examples for more details, and details for more information.
predict_contacts_1y(model, population, age_min = 0, age_max = 100)predict_contacts_1y(model, population, age_min = 0, age_max = 100)
model |
A single fitted model of contact rate (e.g.,
|
population |
a dataframe of age population information, with columns
indicating some lower age limit, and population, (e.g.,
|
age_min |
Age range minimum value. Default: 0 |
age_max |
Age range maximum value, Default: 100 |
Prediction features are added using add_modelling_features().
These features include the population distribution of contact ages,
fraction of population in each age group that attend school/work as well
as the offset according to the settings on all combinations of the
participant & contact ages.
Data frame with four columns: age_from, age_to, contacts, and
se_contacts. This contains the participant & contact ages from the
minimum and maximum ages provided along with the predicted rate of
contacts and standard error around the prediction.
fairfield <- abs_age_lga("Fairfield (C)") fairfield # predict the contact rates in 1 year blocks to Fairfield data fairfield_contacts_1 <- predict_contacts_1y( model = polymod_setting_models$home, population = fairfield, age_min = 0, age_max = 2 )fairfield <- abs_age_lga("Fairfield (C)") fairfield # predict the contact rates in 1 year blocks to Fairfield data fairfield_contacts_1 <- predict_contacts_1y( model = polymod_setting_models$home, population = fairfield, age_min = 0, age_max = 2 )
Predict contact rate for each setting. Note that this function is
parallelisable with future, and will be impacted by any future plans
provided.
predict_setting_contacts( population, contact_model, age_breaks, per_capita_household_size = NULL, model_per_capita_household_size = get_polymod_per_capita_household_size() )predict_setting_contacts( population, contact_model, age_breaks, per_capita_household_size = NULL, model_per_capita_household_size = get_polymod_per_capita_household_size() )
population |
population |
contact_model |
contact_model |
age_breaks |
age_breaks |
per_capita_household_size |
Optional (defaults to NULL). When set, it
adjusts the household contact matrix by some per capita household size.
To set it, provide a single number, the per capita household size. More
information is provided below in Details. See
|
model_per_capita_household_size |
modelled per capita household size.
Default values for this are from
|
We use Per-capita household size instead of mean household size. Per-capita household size is different to mean household size, as the household size averaged over people in the population, not over households, so larger households get upweighted. It is calculated by taking a distribution of the number of households of each size in a population, multiplying the size by the household by the household count to get the number of people with that size of household, and computing the population-weighted average of household sizes. We use per-capita household size as it is a more accurate reflection of the average number of household members a person in the population can have contact with.
List of setting matrices
Nicholas Tierney
# don't run as it takes too long to fit ## Not run: fairfield <- abs_age_lga("Fairfield (C)") fairfield age_break_0_85_plus <- c(seq(0, 85, by = 5), Inf) polymod_contact_data <- get_polymod_setting_data() polymod_survey_data <- get_polymod_population() setting_models <- fit_setting_contacts( contact_data_list = polymod_contact_data, population = polymod_survey_data ) synthetic_settings_5y_fairfield <- predict_setting_contacts( population = fairfield, contact_model = setting_models, age_breaks = age_break_0_85_plus ) fairfield_hh_size <- get_abs_per_capita_household_size(lga = "Fairfield (C)") fairfield_hh_size synthetic_settings_5y_fairfield_hh <- predict_setting_contacts( population = fairfield, contact_model = setting_models, age_breaks = age_break_0_85_plus, per_capita_household_size = fairfield_hh_size ) ## End(Not run)# don't run as it takes too long to fit ## Not run: fairfield <- abs_age_lga("Fairfield (C)") fairfield age_break_0_85_plus <- c(seq(0, 85, by = 5), Inf) polymod_contact_data <- get_polymod_setting_data() polymod_survey_data <- get_polymod_population() setting_models <- fit_setting_contacts( contact_data_list = polymod_contact_data, population = polymod_survey_data ) synthetic_settings_5y_fairfield <- predict_setting_contacts( population = fairfield, contact_model = setting_models, age_breaks = age_break_0_85_plus ) fairfield_hh_size <- get_abs_per_capita_household_size(lga = "Fairfield (C)") fairfield_hh_size synthetic_settings_5y_fairfield_hh <- predict_setting_contacts( population = fairfield, contact_model = setting_models, age_breaks = age_break_0_85_plus, per_capita_household_size = fairfield_hh_size ) ## End(Not run)
Helper function to convert predictions of contact rates in data frames to matrix format with the survey participant age groups as columns and contact age groups as rows.
predictions_to_matrix(contact_predictions, ...)predictions_to_matrix(contact_predictions, ...)
contact_predictions |
data frame with columns |
... |
extra arguments |
Square matrix with the unique age groups from age_group_from/to
in the rows and columns and contacts as the values.
fairfield <- abs_age_lga("Fairfield (C)") # We can convert the predictions into a matrix fairfield_school_contacts <- predict_contacts( model = polymod_setting_models$school, population = fairfield, age_breaks = c(0, 5, 10, 15, Inf) ) fairfield_school_contacts # convert them back to a matrix predictions_to_matrix(fairfield_school_contacts)fairfield <- abs_age_lga("Fairfield (C)") # We can convert the predictions into a matrix fairfield_school_contacts <- predict_contacts( model = polymod_setting_models$school, population = fairfield, age_breaks = c(0, 5, 10, 15, Inf) ) fairfield_school_contacts # convert them back to a matrix predictions_to_matrix(fairfield_school_contacts)
Contact matrices as calculated by Prem. et al. PLoS Computational Biology. DOI: 10.1371/journal.pcbi.1005697
prem_germany_contact_matricesprem_germany_contact_matrices
A list with 5 elements:
A 16x16 matrix containing the number of home contacts, by 5 year age group
A 16x16 matrix containing the number of workplace contacts, by 5 year age group
A 16x16 matrix containing the number of school contacts, by 5 year age group
A 16x16 matrix containing the number of other contacts, by 5 year age group
All age groups are 5 year age bands, from 0 to 80.
https://journals.plos.org/ploscompbiol/article?id=10.1371/journal.pcbi.1005697
Get raw eigvenvalue from NGM matrix
raw_eigenvalue(list_matrix)raw_eigenvalue(list_matrix)
list_matrix |
object of class |
raw eigenvalue
# examples not run as they take a long time ## Not run: perth <- abs_age_lga("Perth (C)") perth_contact <- extrapolate_polymod(perth) perth_ngm <- generate_ngm( perth_contact, age_breaks = c(seq(0, 85, by = 5), Inf) ) raw_eigenvalue(perth_ngm) ## End(Not run)# examples not run as they take a long time ## Not run: perth <- abs_age_lga("Perth (C)") perth_contact <- extrapolate_polymod(perth) perth_ngm <- generate_ngm( perth_contact, age_breaks = c(seq(0, 85, by = 5), Inf) ) raw_eigenvalue(perth_ngm) ## End(Not run)
This value is scaling <- R_target / R_raw, where R_target is the target
R value provided to the NGM, and R_raw is the raw eigenvalue.
scaling(list_matrix)scaling(list_matrix)
list_matrix |
object of class |
scaling
# examples not run as they take a long time ## Not run: perth <- abs_age_lga("Perth (C)") perth_contact <- extrapolate_polymod(perth) perth_ngm <- generate_ngm( perth_contact, age_breaks = c(seq(0, 85, by = 5), Inf) ) raw_eigenvalue(perth_ngm) scaling(perth_ngm) ## End(Not run)# examples not run as they take a long time ## Not run: perth <- abs_age_lga("Perth (C)") perth_contact <- extrapolate_polymod(perth) perth_ngm <- generate_ngm( perth_contact, age_breaks = c(seq(0, 85, by = 5), Inf) ) raw_eigenvalue(perth_ngm) scaling(perth_ngm) ## End(Not run)
Helper function to create your own setting prediction matrix, which you
may want to use in generate_ngm, or autoplot. This class is the
output of functions like extrapolate_polymod, and
predict_setting_contacts. We recommend using this function is only for
advanced users, who are creating their own setting prediction matrix.
setting_prediction_matrix(..., age_breaks)setting_prediction_matrix(..., age_breaks)
... |
list of matrices |
age_breaks |
age breaks - numeric |
setting prediction matrix
age_breaks_0_80_plus <- c(seq(0, 80, by = 10), Inf) one_by_nine <- matrix(1, nrow = 9, ncol = 9) x_example <- setting_prediction_matrix( home = one_by_nine, work = one_by_nine, age_breaks = age_breaks_0_80_plus ) x_example <- setting_prediction_matrix( one_by_nine, one_by_nine, age_breaks = age_breaks_0_80_plus ) x_exampleage_breaks_0_80_plus <- c(seq(0, 80, by = 10), Inf) one_by_nine <- matrix(1, nrow = 9, ncol = 9) x_example <- setting_prediction_matrix( home = one_by_nine, work = one_by_nine, age_breaks = age_breaks_0_80_plus ) x_example <- setting_prediction_matrix( one_by_nine, one_by_nine, age_breaks = age_breaks_0_80_plus ) x_example
see ?get_setting_transmission_matrices for details of how to use these
setting_weightssetting_weights
A named vector of weights relative to home, for home, work, school, and other
Helper function to create your own setting transmission matrix, which you
may want to use in ... or autoplot. This class is the
output of functions like ..., and ... . We recommend using this
function is only for advanced users, who are creating their own
transmission probability matrix.
transmission_probability_matrix(..., age_breaks)transmission_probability_matrix(..., age_breaks)
... |
list of matrices |
age_breaks |
age breaks - numeric |
transmission probability matrix
age_breaks_0_80_plus <- c(seq(0, 80, by = 10), Inf) one_05 <- matrix(0.05, nrow = 9, ncol = 9) x_example <- transmission_probability_matrix( home = one_05, work = one_05, age_breaks = age_breaks_0_80_plus ) x_example <- transmission_probability_matrix( one_05, one_05, age_breaks = age_breaks_0_80_plus ) x_exampleage_breaks_0_80_plus <- c(seq(0, 80, by = 10), Inf) one_05 <- matrix(0.05, nrow = 9, ncol = 9) x_example <- transmission_probability_matrix( home = one_05, work = one_05, age_breaks = age_breaks_0_80_plus ) x_example <- transmission_probability_matrix( one_05, one_05, age_breaks = age_breaks_0_80_plus ) x_example
data frame with information on vaccine coverage, efficacy of acquisition/susceptibility and efficacy of transmission/infectiousness for the ordered age groups from lowest to highest of the next generation matrix.
vaccination_effect_example_datavaccination_effect_example_data
A data frame with 17 rows and 4 variables
character. age bands: 0-4,5-11, 12-15, 16-19, 20-24, etc
example vaccination coverage, between 0-1
example acquisition coverage, between 0-1
example transmission coverage, between 0-1