Phxnlme (version 1.0.0)

phxvpc.plot: Create visual predictive check plot.

Description

This function is used to create a VPC plot using the output from the phxvpc.sim function. The function reads in the output files created by phxvpc.sim and creates a plot. The dependent variable, independent variable and stratification variable are automatically determined from the phxvpc.sim output files.

Usage

phxvpc.plot(vpcpath="", xlab=NULL, ylab=NULL, xlab.cex=1.3, ylab.cex=1.3, x.cex=1.3, y.cex=1.3, main.title=NULL, main.cex=1.3, xlim=NULL, ylim=NULL, obs.pt=FALSE, obs.pch=16, logY=FALSE, Q.obs.line=TRUE, Q.pred.line=TRUE, CI.Q.pred="area", CI.Q.pred.area1="pink", CI.Q.pred.area2="grey", ppp=4, legend=T, result.path=NULL, pred.corr=FALSE, data.obs=NULL, data.Q.obs=NULL, data.Q.pred=NULL, data.Q.CI.pred=NULL)

Arguments

vpcpath

A system directory where vpc simulation results are stored. It should be the same directory as specified in the vpcpath argument in phxvpc.sim

xlab

A title for the x-axis.

ylab

A title for the y-axis.

xlab.cex

Font size for the x-axis title.

ylab.cex

Font size for the y-axis title.

x.cex

Font size for the x-axis label.

y.cex

Font size for the y-axis label.

main.title

A title for the vpc plot.

main.cex

Font size for the vpc plot main title.

xlim

A numerical vector specifying x-axis limits

ylim

A numerical vector specifying y-axis limits

obs.pt

A logical value indicating whether observed data points will be presented. Default is FALSE.

obs.pch

A numerical value specifying the symbol to use when plotting observation data points.

logY

A logical value indicating whether y-axis should be log transformed. Default is FALSE.

Q.obs.line

A logical value indicating whether lines for observation percentiles will be presented. Default is TRUE.

Q.pred.line

A logical value indicating whether lines for prediction percentiles will be presented. Default is TRUE.

CI.Q.pred

A string of either "lines" or "area" (default) specifying whether the confidence intervals of prediction percentiles (as lines or a shaded area) should be added to the plot. NULL means no confidence intervals for prediction percentiles.

CI.Q.pred.area1

A string specifying the color of the shaded area of the confidence intervals for 50th prediction percentile.

CI.Q.pred.area2

A string specifying the color of the shaded area of the confidence intervals for 5th and 95th prediction percentiles.

ppp

Panel per page. Either 1 or 4 (default) specifying the number of panels per page when multiple panels are generated as a result of stratification.

legend

A logical value indicating whether figure legend will be presented. Default is TRUE.

result.path

A system directory where vpc plot will be stored. If NULL (default), it will be set as "../Results", in which ".." is the parent folder of the vpcpath folder.

pred.corr

Predition corrected VPC is not yet implemented in this function.

data.obs

Optional. Default is NULL. Observed data points.

data.Q.obs

Optional. Default is NULL. Data of bservation quantiles.

data.Q.pred

Optional. Default is NULL. Data of predicted quantiles.

data.Q.CI.pred

Optional. Default is NULL. Data of confidence intervals for predicted quantiles.

Value

A plot or a list of plots.

Details

phxvpc.sim must be first executed. The vpc plot will be a pdf file with the same name as the folder that contains phxvpc.sim output. The pdf plot will be stored in the folder specified by the result.path argument.

Examples

## Note: before plotting, first run model fit using phxnlme, 
## next perform VPC simulation using either phxvpc.sim or simmodel.
#setwd("C:/Program Files (x86)/Pharsight/Phoenix/application/Examples/NLME Command Line/Model 3")
#phxvpc.plot(vpcpath="vpc_1")
## or
#phxvpc.plot(vpcpath="C:/Program Files (x86)/Pharsight/Phoenix
#/application/Examples/NLME Command Line/Model 3/vpc_1")

## without showing lines for percentiles of the observation data points
#phxvpc.plot(vpcpath="vpc_1", Q.obs.line=F)

## showing lines for predicted percentiles
#phxvpc.plot(vpcpath="vpc_1", Q.pred.line=T)

## using lines instead of shaded area to indicate the confidence
## intervals for the predicted percentiles
#phxvpc.plot(vpcpath="vpc_1", CI.Q.pred="lines")

## changing color of shaded area
#phxvpc.plot(vpcpath="vpc_1", CI.Q.pred="area", CI.Q.pred.area1="green", CI.Q.pred.area2="yellow")

## changing x-axis limits
#phxvpc.plot(vpcpath="vpc_1", xlim=c(0,10))