prevalenceApprox: Estimates prevalences by age from values by age brackets

In patrickaubert/healthexpectancies: Functions to calculate Health and Disability-free Life Expectancies (DFLE) using Sullivan method

Estimates prevalences by age from values by age brackets

Description

Given a vector of prevalences by age brackets and the vector of age cuts (which defines the age brackets), the function returns a vector of prevalences at all ages. The calculation minimises the sum of squares of second-differences of prevalences by age, under the constraint that average prevalences by age brackets (weight according to the 'weights' vector, usually the vector of population size at each age) are equal to the 'prevalence' input vector.

Usage

prevalenceApprox(
  prevalence,
  agecuts,
  agemin,
  agemax,
  weight = rep(1, (agemax - agemin + 1)),
  categories = c("")
)

Arguments

prevalence	a vector with observed prevalences by age bracket
agecuts	a vector with age defining the age brackets (minimum age in each age bracket)
agemin	minimum age in the output vector
agemax	maximum age in the output vector
weight	a vector of weights for each age
categories	a vector of names of variables of the 'prevalence' dataframe, representing categories (optional)

Details

'prevalence' and 'weight' can also be given as dataframes containing sex, year and/or categories variables. In this case, the output is a dataframe and prevalences are smoothed for all sex, years and categories.

If 'categories' is provided (a vector of variable names in 'prevalence' and 'weights'), smoothing is performed for each separate values of category variables.

Note : Second-differences rather than first-differences are used in the minimisation function, since prevalences according to age are usually parabolic.

Value

a vector with prevalences according to polynomial approximation

Examples

prevalenceApprox(prevalence = (FRDreesAPA2017 %>% filter(sex=="female",typepresta=="APA à domicile"))$prevalence, agecuts=c(seq(60,95,5)), agemin=60, agemax=99, weight=(FRInseePopulation %>% filter(sex=="F",year==2018,age0101>=60,geo=="france") %>% arrange(age0101))$popx)
prevalenceApprox(prevalence = (FRDreesAPA %>% filter(year==2018,sex=="male",typepresta=="APA à domicile"))$prevalence, agecuts=c(seq(60,90,5)), agemin=60, agemax=99, weight=(FRInseePopulation %>% filter(sex=="M",year==2019,age0101>=60,geo=="france") %>% arrange(age0101))$popx)
prevalenceApprox(prevalence = FRDreesAPA2017, agecuts=c(seq(60,95,5)), agemin=60, agemax=99, weight=(FRInseePopulation %>% filter(year==2018,age0101>=60,geo=="france") %>% mutate(sex=recode(sex,"F"="female","M"="male")) %>% rename(weight=popx) %>% arrange(sex,age0101)), categories = c("typepresta"))

patrickaubert/healthexpectancies documentation built on Feb. 5, 2024, 10 a.m.