SIR: SIR test problem

In hetGP: Heteroskedastic Gaussian Process Modeling and Design under Replication

SIR test problem

Description

Epidemiology problem, initial and rescaled to [0,1]^2 versions.

Usage

sirEval(x)

sirSimulate(S0 = 1990, I0 = 10, M = S0 + I0, beta = 0.75, gamma = 0.5, imm = 0)

Arguments

x	vector of size two
S0	initial nunber of susceptibles
I0	initial number of infected
M	total population
beta, gamma, imm	control rates

References

R. Hu, M. Ludkovski (2017), Sequential Design for Ranking Response Surfaces, SIAM/ASA Journal on Uncertainty Quantification, 5(1), 212-239.

Examples

## SIR test problem illustration
ngrid <- 10 # increase
xgrid <- seq(0, 1, length.out = ngrid)
Xgrid <- as.matrix(expand.grid(xgrid, xgrid))

nrep <- 5 # increase
X <- Xgrid[rep(1:nrow(Xgrid), nrep),]
Y <- apply(X, 1, sirEval)
dataSIR <- find_reps(X, Y)
filled.contour(xgrid, xgrid, matrix(lapply(dataSIR$Zlist, sd), ngrid),
               xlab = "Susceptibles", ylab = "Infecteds", color.palette = terrain.colors)

hetGP documentation built on Sept. 11, 2024, 6:56 p.m.