contact-fitter-loglinear: Fitting Contact Data onto Serological Data - Age-Dependent...

In TeaKov/serostat: Modeling Infectious Disease Parameters Based on Serological and Social Contact Data R Package

Description

Whereas in contact.fitter.location refinement was based on stratification, we now argue that the proportionality factor “q” might depend on several characteristics related to susceptibility and infectiousness, which could be ethnic-, climate-, disease-, or age-specific.

Usage

contact.fitter.loglinear(a, y, rij, int, muy, N, D, Lmax, A, plots, startpar)

Arguments

a	Numeric vector of age categories.
y	The response variable (binary, 1=past infection, 0=otherwise).
rij	The smoothed contact matrix.
int	Specifies if there is a third gamma variable or not.
muy	Mortality function.
N	Population size.
D	Mean infectious period.
Lmax	Maximum age.
A	Age of loss of maternal immunity (0<A<1).
plots	Generate plots during fitting process.
startpar	Starting values for the “nlm” method used in this function.

See Also

pwcrate, waifw.6parms

Examples

ND <- c(489,47,29,21,12,12,16,15,15,6,6,14,17,19,17,23,34,33,62,71,68,68,78,71,71,96,86,83,79,
  80,83,93,126,120,121,132,135,176,161,193,196,218,257,277,331,376,356,435,460,453,535,545,
  576,668,692,759,722,819,939,1015,1051,973,1113,996,940,1074,1252,1367,1468,1541,1661,1838,
  2012,2236,2517,2793,2938,2994,3311,3516,3727,3857,4088,4161,4261,4274,4061,2509,2049,2159,
  2205,2550,2330,1992,1569,1242,1000,726,533,996)
PS <- c(118366,117271,114562,113894,116275,118030,116761,117742,119583,119887,118963,119958,
  124637,129143,131030,129724,127187,126433,124377,124883,122201,124482,126459,130129,133897,
  135009,134516,133495,132705,132040,130602,135638,140537,146151,150467,152113,151656,151412,
  153371,158268,162456,167652,164871,161671,162060,159735,160672,157030,153820,151114,148978,
  145929,142374,141215,135525,135968,134692,135991,134291,134131,113024,112198,105880,92772,
  84462,93787,100820,101866,97208,94145,92451,93027,91640,93593,91933,89900,81718,77891,73104,
  70082,67057,62178,57642,51786,47466,42065,28004,17186,14492,13838,13957,13358,10442,8063,5604,
  4289,2843,2068,1368,2146)
AGE<-c(0:(length(ND)-1))

estimL <- estimateLifeExpectancy(ND, PS, AGE)

# Using Belgian B19 data
data("VZV_B19_BE_0103")
VZV_B19_BE_0103 <- VZV_B19_BE_0103[!is.na(VZV_B19_BE_0103$parvores)&!is.na(VZV_B19_BE_0103$age)&
  VZV_B19_BE_0103$age<70&VZV_B19_BE_0103$age>=1,]
VZV_B19_BE_0103 <- VZV_B19_BE_0103[order(VZV_B19_BE_0103$age),]

y <- VZV_B19_BE_0103$parvores
a <- VZV_B19_BE_0103$age

# Mean duration of infectiousness
D <- 6/365

# Maximum life (if type mortality this is the life expectancy)
Lmax<-100

# Age of loss of maternal immunity (0<A<1)
A <- 0.5

# Mortality function
My <- estimL$My[1:Lmax]
muy <- estimL$muy[1:Lmax]

# Population size
N <- sum(PS)

data("scd_close_p4h")

contact.result<-contact.fitter.loglinear(a=a,y=y,rij=scd_close_p4h,int=FALSE,
  muy=muy,N=N,D=D,Lmax=85,A=A,plots="TRUE",startpar=c(-2.3,0,0))
c(contact.result$qhat,contact.result$R0,contact.result$aic)
contact.result$qhat/sqrt(diag(solve(contact.result$qhess)))

contact.result<-contact.fitter.loglinear(a=a,y=y,int=TRUE,rij=scd_close_p4h,
  muy=muy,N=N,D=D,Lmax=85,A=A,plots="TRUE",startpar=c(-2.3,0,0))
c(contact.result$qhat,contact.result$R0,contact.result$aic)
contact.result$qhat/sqrt(diag(solve(contact.result$qhess)))

TeaKov/serostat documentation built on May 21, 2019, 1:21 p.m.