globalval.fun: Perform a global validation analysis for a NHPP In NHPoisson: Modelling and Validation of Non Homogeneous Poisson Processes

Description

This function performs a thorough validation analysis for a fitted NHPP. It calculates the (generalized) uniform and the raw (or scaled) residuals, performs residual plots for the uniform residuals, and time residual and lurking variable plots for the raw or scaled residuals. It also plots the fitted and empirical estimations of the NHPP intensity. Optionally, it also performs a residual QQplot.

Usage

 1 2 3 4 5 6 globalval.fun(mlePP, lint = NULL, nint = NULL, Xvar = NULL, namXvar = NULL, Xvart = NULL, namXvart = NULL, h = NULL, typeRes = NULL, typeResLV="Pearson",typeI = "Disjoint", nsim = 100, clevel = 0.95, resqqplot = FALSE, nintLP = 100, tit = "", flow = 0.5, addlow = FALSE, histWgraph=TRUE,plotDisp=c(2,2), indgraph = FALSE, scax = NULL, scay = NULL, legcex = 0.5, cores = 1, xlegend = "topleft", fixed.seed=NULL) 

Arguments

 mlePP An object of class mlePP-class; usually, the output from fitPP.fun. lint Length of the intervals used to calculate the residuals. nint Number of intervals used to calculate the residuals. Intervals of equal length are considered. Only used if typeI="Disjoint". In that case, only one of the arguments lint or nint must be specified. Xvar Optional. Matrix of the lurking variables (each column is a variable). namXvar Optional. Vector of names of the variables in Xvar. Xvart Optional. Matrix of the variables for the residual plots (each column is a variable). A time plot is performed in all the cases. namXvart Optional. Vector of names of the variables in Xvart. h Optional. Weight function to calculate the scaled residuals. By default, Pearson residuals with h(t)=1/√{\hat λ(t)} are calculated. This function is used to calculate both the scaled residuals and the residuals for the lurking variables (except if typeResLV="Raw"). typeRes Optional. Label indicating the type of scaled residuals. By default, Pearson residuals are calculated and label is "Pearson". typeResLV Label indicating the type of residuals ("Raw" or any type of scaled residuals such as "Pearson") to calculate the residuals for the lurking variable plots. typeI Label indicating the type ("Overlapping" or "Disjoint") of intervals used to calculate the residuals. clevel Confidence level of the residual envelopes. resqqplot Logical flag. It is is TRUE, a residual qqplot is carried out. nsim Number of simulations for the residual qqplot. nintLP Number of levels considered in the lurking variables. It is used as argument nint in the call of the function graphResCov.fun. tit Character string. A title for the plot. flow Argument f for the lowess smoother of the raw (or scaled) residual plots, see lowess. addlow Logical flag. If it is TRUE, a lowess is added in the residual plots. histWgraph Logical flag. If it is TRUE, a new graphical device is opened with the option record=TRUE, so that the history of all plots is recorded in the new device. This option may not work on some platforms; for example, RStudio does not allow the user to open new graphical devices. plotDisp A vector of the form c(nr, nc). The residual plots in graphresU.fun, graphres.fun and graphResCov.fun will be drawn in a nr\timesnc layout. It is used as argument mfrow in par. By default, a 2 \times 2 layout is used. indgraph Logical flag. If it is TRUE, the validation plots (except the residual versus variables plots) in graphresU.fun are carried out in four 1 \times 1 layouts. By default, a 2 \times 2 layout is used. scax Optional. Vector of two values indicating the range of values for the x-axis in the fitted and empirical rate plot. An adequate range is selected by default. scay Optional. Vector of two values indicating the range of values for the x-axis in the fitted and empirical rate plot. An adequate range is selected by default. legcex cex argument for the legend in the residual time plots (see par for details). cores Optional. Number of cores of the computer to be used in the calculations. Default: one core is used. xlegend Argument xlegend used in the call of the function graphrate.fun; see that function for details. fixed.seed An integer or NULL. It is the argument for resQQplot.fun.

Details

If typeI="Overlapping", argument lint is compulsory. If typeI="Disjoint", only one of the arguments lint or nlint must be specified.

Value

A list with the same elements that CalcRes.fun or CalcResD.fun (depending on the value of the argument typeI).

References

Cebrian, A.C., Abaurrea, J. and Asin, J. (2015). NHPoisson: An R Package for Fitting and Validating Nonhomogeneous Poisson Processes. Journal of Statistical Software, 64(6), 1-24.

graphres.fun, graphrate.fun, resQQplot.fun, graphResCov.fun, graphresU.fun
  1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 data(BarTxTn) covB<-cbind(cos(2*pi*BarTxTn$dia/365), sin(2*pi*BarTxTn$dia/365), BarTxTn$TTx,BarTxTn$Txm31,BarTxTn$Txm31**2) modB<-fitPP.fun(tind=TRUE,covariates=covB, POTob=list(T=BarTxTn$Tx, thres=318), tit="BAR Tx; cos, sin, TTx, Txm31, Txm31**2", start=list(b0=-100,b1=1,b2=10,b3=0,b4=0,b5=0),CIty="Transf",modCI=TRUE, modSim=TRUE,dplot=FALSE) #Since only one graphical device is opened and the argument histWgraph is TRUE #by default, the different plots can be scrolled up and down with the "Page Up" #and "Page Down" keys. aux<-globalval.fun(mlePP=modB,lint=153, typeI="Disjoint", typeRes="Raw",typeResLV="Raw", resqqplot=FALSE) #If typeRes and typeResLV are not specified, Pearson residuals are calculated #by default. aux<-globalval.fun(mlePP=modB,lint=153, typeI="Disjoint", resqqplot=FALSE)