plot.densLPS function

Plot the density estimate in a densLPS.object

Plot the density estimate obtained by densityLPS from censored data with given mean and variance.

## S3 method for class 'densLPS'
plot(x,
       xlim=range(fit$bins),breaks=NULL,hist=FALSE,histRC=FALSE,
       xlab="",ylab="Density",main="",...)

Arguments

• x: a densLPS.object.
• xlim: interval of values where the density should be plotted.
• breaks: (Optional) breaks for the histogram of the observed residuals.
• hist: Logical (Default: FALSE) indicating whether the histogram of the (pseudo-) data should be plotted with the estimated density.
• histRC: Logical (Default: FALSE) indicating whether the histogram of the right-censored residuals should be highlighted.
• xlab: Optional label for the x-axis (Defaut: empty).
• ylab: Optional label for the y-axis (Default: "Density").
• main: Plot main title (Default: "").
• ...: Optional additional plot parameters.

Returns

No returned value (just plots).

Examples

require(DALSM)

## Example 1: density estimation from simulated IC data
n = 500 ## Sample size
x = 3 + rgamma(n,10,2) ## Exact generated data
width = runif(n,1,3) ## Width of the IC data (mean width = 2)
w = runif(n) ## Positioning of the exact data within the interval
xmat = cbind(x-w*width,x+(1-w)*width) ## Generated IC data
head(xmat)
obj.data = Dens1d(xmat,ymin=0) ## Prepare the data for estimation
## Density estimation with fixed mean and variance
obj = densityLPS(obj.data,Mean0=3+10/2,Var0=10/4)
plot(obj, hist=TRUE) ## Histogram of the pseudo-data with the density estimate
curve(dgamma(x-3,10,2), ## ... compared to the true density (in red)
      add=TRUE,col="red",lwd=2,lty=2)
legend("topright",col=c("black","red","grey"),lwd=c(2,2,10),lty=c(1,2,1),
       legend=c("Fitted density","True density","Pseudo-data"),bty="n")
print(obj) ## ... with summary statistics

## Example 2: estimation of the error density in a DALSM model
data(DALSM_IncomeData)
resp = DALSM_IncomeData[,1:2]
fit = DALSM(y=resp,
            formula1 = ~twoincomes+s(age)+s(eduyrs),
            formula2 = ~twoincomes+s(age)+s(eduyrs),
            data = DALSM_IncomeData)
plot(fit$derr, hist=TRUE)  ## Plot the estimated error density
legend("topright",col=c("black","grey"),lwd=c(2,10),lty=c(1,1),
       legend=c("Estimated error density","Pseudo-residuals"),bty="n")
print(fit$derr) ## ... and provide summary statistics for it

References

Lambert, P. (2021). Fast Bayesian inference using Laplace approximations in nonparametric double additive location-scale models with right- and interval-censored data. Computational Statistics and Data Analysis, 161: 107250. doi:10.1016/j.csda.2021.107250

See Also

densLPS.object, print.densLPS, densityLPS.

Author(s)

Philippe Lambert p.lambert@uliege.be