vpcfig2: Visual Predictive Check (VPC) based on Perl-speaks-NONMEM...

In nonmem2R: Loading NONMEM Output Files with Functions for Visual Predictive Checks (VPC) and Goodness of Fit (GOF) Plots

Visual Predictive Check (VPC) based on Perl-speaks-NONMEM (PsN) generated VPC files (ggplot2-version).

Description

This function creates VPC using output files from the vpc command in Pearl Speaks NONMEM (PsN). Graphs are generated using the ggplot2 package and the return object is an ggplot class and can be further modified, see details below and documentation for ggplot2 for further details.

Usage

vpcfig2(vpcdir = NULL, vpctab = NULL, vpcresult = NULL,
  use.model.path = TRUE, strata.names = NULL, strata.subset = NULL,
  median.only = FALSE, bin.idv = c("median", "midpoint"),
  percentile = 10, fy = function(y) {     y }, fx = function(x) {    
  x }, xlab = NULL, ylab = NULL, col.data = "gray20", cex.data = 1,
  pch.data = 19, alpha.data = 0.5, col.line = "grey20",
  lwd.line = 1, lty.line = c(2, 1, 2), col.segm = c("steelblue",
  "gray50", "steelblue"), alpha.segm = 0.6, type = 3,
  panel.height.censored = 0.25, censoring.labels = c("ALQ", "BLQ"),
  ignore.censoring = FALSE, control = GOF.control())

Arguments

vpcdir	Path of directory of the VPC files
vpctab	Path to the vpctab-file
vpcresult	Path to the vpcresult file
use.model.path	Load file from a global defined model library (TRUE=default). If so will look for a global character vector named model.path
strata.names	Character vector for strata names. Must have length equal to number of strata's in vpc files, otherwise ignored.
strata.subset	Vector specifying subset of strata to use. Either a vector of index, e.g. strata.subset=c(1,3), or a character vector naming which strata to use. See details.
median.only	Logical to plot only median and no quantiles for model and data (FALSE=default)
bin.idv	Method for computed idv value for plotting bin summary values. Set to "median" to use median of independent variable in each bin on x-axis (default), or set to "midpoint" to use midpoint of bins on x-axis.
percentile	percentile to use, default=10 will display 10 Percentile argument must match columns included in the vpcresult file.
fy	transformation function for y-axis, default to identity function (f(y)=y).
fx	transformation function for x-axis, default to identity function (f(x)=x)
xlab	label for x-axis, default is IDV as found in vpcresultfile, passed to labs.
ylab	label for y-axis, default is DV as found in vpcresultfile, passed to labs.
col.data	color for data points, passed to geom_point, default="gray20"
cex.data	cex for data points, passed to geom_point, default=1
pch.data	plot symbol (pch) for data points, passed to geom_point, default=19
alpha.data	alpha for plotting of data points, passed to geom_point, default=0.5
col.line	line color for low quartile, median, and high quartile of data, passed to geom_line, default="gray20"
lwd.line	line width (lwd) for low quartile, median, and high quartile of data, passed to geom_line, default=1
lty.line	line type (lty) for low quartile, median, and high quartile of data, passed to geom_line, should be vector of length 3, default=c(2,1,2)
col.segm	Color of CI regions (low, median and high) for model, passed to geom_polygon, should be vector of length 3, default=c("steelblue", "gray50", "steelblue")
alpha.segm	alpha of CI regions (low, median and high) for model, passed to geom_polygon, default=0.6
type	type of VPC plot: 1=display model regions only, 2=as 1 + lines (low, median and high) of data, 3= as 2 + points for data. type=0 can also be used and then no graph is produced but instead a list with 2 dataframes is returned. one for the vpcresult and on for the observed data in the vpctab file.
panel.height.censored	Height for BLQ ALQ panels. Height is relative to height of Un-cenored panel, default=0.25
censoring.labels	Character vector used as labels for panels when VPC have either BLQ or ALQ data Default set to c("ALQ","BLQ"). NOTE: must be character vector of length 2
ignore.censoring	Ignore any censoring results in VPC result file (BLQ and ALQ), defult=FALSE
control	an optional list of control settings. See GOF.control for the names of the settable control values and their effect.

Details

The ggplot2 package is used for creating the VPC and vpcfig2 have functionality for e.g. modify names of strata, order of and subset of strata's, and transform the x- and or y-axis before plotting.

The data used is either specified by the directory of the PsN generated files, or by specifying the file names of both the vpctab-file and the vpcresult file. See examples 1 and 2 below.

Names of strata can be changed with the strata.names argument, if strata.names is NULL the names as specified in the vpcresult file are used.

strata.subset can be used to select a subset of strata, and or to change the order of stratas. See example 3.

Strata.subset is matched with strata.names unless strata.names=NULL. If strata.names=NULL then strata.subset is matched to names as specified in vpcresult file. Default is to use all strata's (strata.subset=NULL).

The fy and fx arguments can be used to alter the scale of data plotted. For example, using fy=function(y){log(y+1)} is a convenient way to get log-scale for y-axis but with an off-set to show any values==0. Proper y tick marks can then be set using the scale_y_continuous ggplot2 function.

Since the returned objest is a ggplot-class object is can be further modified to e.g. to log-scale for y.axis;

vpcfig2(...)+scale_y_log10(),

adding/changing labels & titles;

vpcfig2(...)+labs(y="Modified y-label", title="New title").

Axis-limits are preferably set using the ggplot2 function coord_cartesian(...). This way data points outside the axis-limit are only hidden when plotting.

Stratified VPC's are created with facet_wrap(~strata) as deafult but can be modified to use facet_grid for setting the panel grid.

However faceting must be done with ~strata.

See examples below and documentation for ggplot2 for further details.

Value

ggplot object of VPC plot

Examples

# Get path to the example files included in nonmem2R package
file1 <- system.file("extdata", "vpctab004.dat", package = "nonmem2R")
file2 <- system.file("extdata", "vpc_results.csv", package = "nonmem2R")


# Ex 1, produce VPC with default setting, here specifying both vpctab and vpcresult
vpcfig2(vpctab=file1,vpcresult=file2)

## Not run: 
# Ex 2, produce VPC with default setting, here specifying only directory of vpc files
path1<-gsub("vpctab004.dat","",file1)
vpcfig2(vpcdir=path1)

## End(Not run)

# Ex 3, produce VPC with i) modifies strata names, ii) strata in reverse order, and
#        iii) labels
strata.names<-c("Group A","Group B")
xlab<-"Time after dose (hrs)"
ylab<-"Plasma Conc(mmol/L)"
vpcfig2(vpctab=file1,vpcresult=file2,strata.names=strata.names,strata.subset=2:1,
       xlab=xlab,ylab=ylab)


# Example using the fy argument to transform y-axis setting y-ticks using scale_y_continuous(...)
tmp<-c(1,2,3,4,6)
yticks<-c(0.1,tmp,tmp*10,tmp*100,tmp*1000)
vpcfig2(vpctab=file1,vpcresult=file2,fy=function(y){log(y+1)})+
 scale_y_continuous(breaks=log(yticks+1),labels=yticks,minor_breaks=NULL)

## Not run: 
# Example changing to slog-scale using the ggplot2 function scale_y_log10
vpcfig2(vpctab=file1,vpcresult=file2)+scale_y_log10()

## End(Not run)
# Example changing y-axis label and adding figure title using ggplot2 function labs(...).
vpcfig2(vpctab=file1,vpcresult=file2)+labs(y="Modified y-label", title="New title")

nonmem2R documentation built on April 15, 2022, 1:05 a.m.