PlotPredictiveCheck {MIfuns}R Documentation

Plots results of the ComputePredictiveCheck function

Description

The function PlotPredictiveCheck uses the output list of the ComputePredictiveCheck function to create histograms that compare distribution of study statistics across the simulated studies with the statistic computed for the observed data.

Usage

PlotPredictiveCheck(PCData, Parameters = c("min", "max", "mean",
"median", "Tmin", "Tmax", "AUC"), Summaries = c("min", "max", "mean",
"median", "Q25", "Q75"), plot.hist = TRUE, plot.qq = TRUE,
DrugName = "DV", plotType = "pdf", FileTemplate = "Plots")

Arguments

PCData The output of the ComputePredictiveCheck function
Parameters list of individual parameter names (in quotes) that should be illustrated by plots. These parameter names should be either from SimpleFun or CombFun lists of the ComputePredictiveCheck function call that created PCData.
Summaries list of summary statistics that are used to characterize the distribution of individual parameters. These parameter names should be from the StatFun list of the ComputePredictiveCheck function call that created PCData.
plot.hist equal to TRUE (default value) if histogram plot is requested and FALSE otherwise.
plot.qq equal to TRUE (default value) if QQ-plot is requested and FALSE otherwise.
DrugName Name of the drug (default value is “DV”) that will be used in the plots. Limited control over plot labels and title exist. The function needs to be modified in order to create more precise titles and labels.
plotType Character name (in quotes) that assigns the device for the output plots. Supported names are “wmf”, “pdf”, “ps”, “png”, “fpeg”, and “bmp” that initiate the appropriate R graphic device and create one or several output files. The default value “pdf” is suitable for the exploratory investigation of the data. For report-quality plots, on MSWindows only, “wmf” is recommended. If PlotType is “pdf” or “ps”, one file with all the figures is created. For any other PlotType, one file per plot is created. Extension of the file coincides with the PlotType.
FileTemplate File name template that will be used to create the output file name. If one file with all the plots is created (PlotType is “pdf” or “ps”) then the PlotType extension will be added to the FileTemplate name to create the name of the output file. If one file per plot is created (PlotType is not equal to “pdf” or “ps”) then the three-digit order number will be added to the FileTemplate name. Then the PlotType extension will be added to create the name of the output file.

Value

The output of the program is one or more files with the plots. If plot.hist is TRUE, then for each combination of the elements from the Parameters and Statistics lists, the histogram of simulated values is created; observed value is shown by the vertical line; probability of observing statistics as extreme as observed data is presented in the title of each histogram. If plot.qq is TRUE, then for each element of the Parameters lists, the quantile-quantile plots are created. The plots superimpose quantile-quantile curves of the parameter distributions in each of the simulated studies against the observed parameter distribution. The unit line is provided for the reference.

Author(s)

Developed by Metrum Institute (Bill Knebel, Tim Bergsma, and others).

References

MIfuns package available from Metrum Institute website (http://metruminstitute.org/).


[Package MIfuns version 3.2.9 Index]