twinSIR_intensityplot: Plotting Paths of Infection Intensities for 'twinSIR' Models

In surveillance: Temporal and Spatio-Temporal Modeling and Monitoring of Epidemic Phenomena

Plotting Paths of Infection Intensities for twinSIR Models

Description

intensityplot methods to plot the evolution of the total infection intensity, its epidemic proportion or its endemic proportion over time. The default plot method for objects of class "twinSIR" is just a wrapper for the intensityplot method. The implementation is illustrated in Meyer et al. (2017, Section 4), see vignette("twinSIR").

Usage

## S3 method for class 'twinSIR'
plot(x, which = c("epidemic proportion", "endemic proportion",
     "total intensity"), ...)

## S3 method for class 'twinSIR'
intensityplot(x, which = c("epidemic proportion", "endemic proportion",
              "total intensity"), aggregate = TRUE, theta = NULL,
              plot = TRUE, add = FALSE, rug.opts = list(), ...)

## S3 method for class 'simEpidata'
intensityplot(x, which = c("epidemic proportion", "endemic proportion",
              "total intensity"), aggregate = TRUE, theta = NULL,
              plot = TRUE, add = FALSE, rug.opts = list(), ...)

Arguments

x	an object of class "twinSIR" (fitted model) or "simEpidata" (simulated twinSIR epidemic), respectively.
which	"epidemic proportion", "endemic proportion", or "total intensity". Partial matching is applied. Determines whether to plot the path of the total intensity \lambda(t) or its epidemic or endemic proportions \frac{e(t)}{\lambda(t)} or \frac{h(t)}{\lambda(t)}.
aggregate	logical. Determines whether lines for all individual infection intensities should be drawn (FALSE) or their sum only (TRUE, the default).
theta	numeric vector of model coefficients. If x is of class "twinSIR", then theta = c(alpha, beta), where beta consists of the coefficients of the piecewise constant log-baseline function and the coefficients of the endemic (cox) predictor. If x is of class "simEpidata", then theta = c(alpha, 1, betarest), where 1 refers to the (true) log-baseline used in the simulation and betarest is the vector of the remaining coefficients of the endemic (cox) predictor. The default (NULL) means that the fitted or true parameters, respectively, will be used.
plot	logical indicating if a plot is desired, defaults to TRUE. Otherwise, only the data of the plot will be returned. Especially with aggregate = FALSE and many individuals one might e.g. consider to plot a subset of the individual intensity paths only or do some further calculations/analysis of the infection intensities.
add	logical. If TRUE, paths are added to the current plot, using lines.
rug.opts	either a list of arguments passed to the function rug, or NULL (or NA), in which case no rug will be plotted. By default, the argument ticksize is set to 0.02 and quiet is set to TRUE. Note that the argument x of the rug() function, which contains the locations for the rug is fixed internally and can not be modified. The locations of the rug are the time points of infections.
...	For the plot.twinSIR method, arguments passed to intensityplot.twinSIR. For the intensityplot methods, further graphical parameters passed to the function matplot, e.g. lty, lwd, col, xlab, ylab and main. Note that the matplot arguments x, y, type and add are implicit and can not be specified here.

Value

numeric matrix with the first column "stop" and as many rows as there are "stop" time points in the event history x. The other columns depend on the argument aggregate: if TRUE, there is only one other column named which, which contains the values of which at the respective "stop" time points. Otherwise, if aggregate = FALSE, there is one column for each individual, each of them containing the individual which at the respective "stop" time points.

Author(s)

Sebastian Meyer

References

Meyer, S., Held, L. and Höhle, M. (2017): Spatio-temporal analysis of epidemic phenomena using the R package surveillance. Journal of Statistical Software, 77 (11), 1-55. \Sexpr[results=rd]{tools:::Rd_expr_doi("10.18637/jss.v077.i11")}

See Also

twinSIR for a description of the intensity model, and simulate.twinSIR for the simulation of epidemic data according to a twinSIR specification.

Examples

data("hagelloch")
plot(hagelloch)

# a simplistic twinSIR model
fit <- twinSIR(~ household, data = hagelloch)

# overall total intensity
plot(fit, which = "total")

# overall epidemic proportion
epi <- plot(fit, which = "epidemic", ylim = c(0, 1))
head(epi)
# add overall endemic proportion = 1 - epidemic proportion
ende <- plot(fit, which = "endemic", add = TRUE, col = 2)
legend("topleft", legend = "endemic proportion", lty = 1, col = 2, bty = "n")

# individual intensities
tmp <- plot(fit, which = "total", aggregate = FALSE,
    col = rgb(0, 0, 0, alpha = 0.1),
    main = expression("Individual infection intensities " *
        lambda[i](t) == Y[i](t) %.% (e[i](t) + h[i](t))))
# return value: matrix of individual intensity paths
str(tmp)

# plot intensity path only for individuals 3 and 99
matplot(x = tmp[,1], y = tmp[,1+c(3,99)], type = "S",
        ylab = "Force of infection", xlab = "time",
        main = expression("Paths of the infection intensities " *
                          lambda[3](t) * " and " * lambda[99](t)))
legend("topright", legend = paste("Individual", c(3,99)),
       col = 1:2, lty = 1:2)

surveillance documentation built on Nov. 4, 2024, 3 a.m.