w.matrix: Creates a weight matrix from (1) immunization information for...

In vacem: Vaccination Activities Coverage Estimation Model

Description

Creates a weight matrix from (1) immunization information for a set of individuals, e.g. observations derived from a Demographic and Health Survey (DHS), (2) a set of vaccination activity descriptions, and (3) a cumulative distribution function (CDF).

Usage

w.matrix(obs, activities, cdf.fun, ...)

Arguments

obs	a data frame with one row per observation and columns for date of observation and age of individual on that date; <i>Note: all ages should be in months</i>
activities	a data frame with one row per activity and columns providing the activity's date and the @c is.SIA flag indicating whether the activity is a Supplemental Immunization Activity (SIA) or routine vaccination.
cdf.fun	a cumulative distribution function (CDF) used to calculate the probability of routine vaccination during each activity as a function of age, e.g. @code constant.post.9mo .
...	additional arguments to the @p cdf.fun, e.g. @c lambda

Details

The @c w.matrix function loops through all the provided activities and determines weights for all observations. These weights give the probability of an individual having the "opportunity" to be vaccinated during a year's routine activities.

Value

a matrix with one row per observation and one column per activity; matrix cells contain an "opportunity" weight which is 0 for SIA campaigns and calculated as using the @p cdf.fun for routine vaccination activities

@see nine.month.pointmass @see constant.post.9mo

Note

For all Supplemental Immunization Activities (SIAs), i.e. campaigns, the routine vaccination opportunity weight defaults to zero (0).

For each routine vaccination activity, the opportunity weight is based on the individual's age at the start of the activity and the length of exposure to that year's activity (12 months most years, but truncated in the survey year). Specifically, the weight is calculated as: @f[ cdf.fun( xij + lj ) - cdf.fun( xij ) @f] where @c xij is the age of individual @c i at the start of year @c j and @c lj is the length of exposure to routine vaccination in year @c j.

For length of the exposure to the routine vaccination, @c lj, there are three possible cases: @li a. if survey date is before activity year, then the individual is not "exposed" to this routine vaccination and hence @f$ lj = 0 months @f$ @li b. if survey date is during activity year, then the individual is "exposed" only up until the observation date and hence @f$ lj = (obs$date - activities$date) @f$ @li c. if survey date is after activity year, then the individual is "exposed" to this routine vaccination for the entire year and hence @f$ lj == 12 months @f$

Note

For routine vaccinations, the date provided (i.e. @p activities$date) is used as start of the activity year and hence is typically given as January 1st of that year, e.g. @c 2012-01-01.

vacem documentation built on April 14, 2017, 7:28 p.m.