Skip to contents

Plot settings

Source: vignettes/plot-settings.Rmd
plot-settings.Rmd
require(ospsuite.reportingengine)
#> Loading required package: ospsuite.reportingengine
#> Loading required package: ospsuite
#> Loading required package: rSharp
#> 
#> Attaching package: 'ospsuite'
#> The following object is masked from 'package:base':
#> 
#>     %||%
#> Loading required package: tlf

Overview

This vignette provides some guidance to define properties and settings for the reporting engine.

These properties can at 2 levels:

  • Global or workflow plot configuration, affecting every plot of the workflow output
  • Local or task plot configuration, affecting specific plots within a task

How to set global workflow properties

Reporting Engine global setting names are listed in the enum reSettingsNames. Each global setting can be reviewed using the function getRESettings().

It is possible to update, save and reuse most of the reporting engine default settings using the functions available in the section Default Settings.

Note that the function resetRESettingsToDefault() allows you to reset to initial default values defined by the reporting engine.

Notion of theme

A theme is an object of R6 class Theme which defines the default properties used to build every plot in tlf package. More information can be found directly within the documentation of the the tlf package (vignettes “theme-maker” and “plot-configuration”).

The best workflow to set a Theme object is to use json files. Templates of json files are available in the tlf package at the location defined below:

# Get the path of the Json template
jsonFile <- system.file("extdata", "re-theme.json",
  package = "ospsuite.reportingengine"
)

# Load the theme from the json file
reTheme <- tlf::loadThemeFromJson(jsonFile)

Modifications of the theme properties can be done directly by updating the properties of the Theme object as shown below:

# Overwrite the current background fill
reTheme$background$panel$fill <- "white"
reTheme$background$panel$fill
#> [1] "white"

To save the updated Theme object, the function saveThemeToJson provided by tlf package can be used.

# Save your updated theme in "my-new-theme.json"
tlf::saveThemeToJson("my-new-theme.json", theme = reTheme)

In addition, the tlf package provides a Shiny app to define and save your own theme with a few clicks: tlf::runThemeMaker().

Then, the most important step is to define the Theme object as the current default for the workflow. In ospsuite.reportingengine, this can be either achieved using the function setDefaultTheme with a Theme object as shown in the example below with reTheme.

The properties defined by the Theme object will be used by default if no specific plot configurations are provided for the workflow plots.

# Set reTheme as current default Theme
setDefaultTheme(reTheme)

Caution: loading a workflow re-initialize the default theme. The function setDefaultTheme() needs to be called after initialization of the workflow.

Additionally, Workflow objects have a theme input argument available in method $new() to use the input theme instead of the reporting engine default.

Exporting plots in a specific format

One important global setting is the format of the figure files (e.g. png, jpg or pdf) as well as the quality of the figures. By default, the figures are saved as png files with a size of 20 x 15 cm. The format settings can be managed using the function setDefaultPlotFormat() as illustrated below, input arguments are format, width, height, their units and the plot resolution dpi.

setDefaultPlotFormat(format = "png")

Watermark

Watermark corresponds to text in the background of plots. Watermark font properties can be set in the Theme object: reTheme$fonts$watermark.

Regarding the content of the watermark, the default content available in the Theme object (i.e. at reTheme$background$watermark in the example) is overwritten by the Workflow objects:

  • If a system is validated, no text is printed in background by default.
  • If a system is NOT validated, the text “preliminary results” is printed in background by default.

As a consequence, setting the watermark content needs to be performed directly with the Workflow objects: either when initializing the workflow using the input argument watermark (e.g. MeanModelWorkflow$new(..., watermark = "vignette example")); or using the workflow method $setWatermark(watermark) (e.g. myWorkflow$setWatermark("vignette example")).

The workflow method $getWatermark() prints the watermark text which will be in the background of the figures output by the workflow (e.g. myWorkflow$getWatermark()).

Note: The NULL input argument for watermark leads to using the default watermark feature.

How to set task specific properties

Each task of Workflow objects includes a field named settings. This optional field of class TaskSettings can be updated and will overwrite the default behavior of a task.

Two usual subfields of settings are: plotConfigurations and bins. Other settings can also be provided and will be detailed in the next sections.

  • plotConfigurations is a list of PlotConfiguration objects from the tlf package. PlotConfiguration is an R6 class in which plot properties are stored and used to create plots with a desired configuration. More information can be found on this notion within the tlf vignette “plot-configuration”. Providing a user-defined PlotConfiguration object to this input list will overwrite the default configuration and will be used to generate the task corresponding task plot(s).
  • bins can be a number, a vector or a function which will be used to bin histograms.

The next section will provide the names of the plots and options that can be accessed through the field settings.

Time profiles and residuals

Regardless of the workflow subclass (MeanModelWorkflow or PopulationWorkflow), the names in the plotConfigurations list available for the “plotTimeProfilesAndResiduals” task are the following:

  • “timeProfile”: for time profile plots in linear and log scale
  • “obsVsPred”: for observed vs simulated data in linear and log scale
  • “resVsPred”: for residuals vs simulated data plots
  • “resVsTime”: for residuals vs time plots
  • “resHisto”: for histograms of residuals
  • “resQQPlot”: for qq-plots of residuals
  • “histogram”: for histograms of residuals across all simulations
  • “qqPlot”: for qq-plots of residuals across all simulations

Since histograms are performed by the task, bins can also be defined.

Displayed names and units are usually managed through the SimulationSet and Output objects. The vignette “goodness-of-fit” provides more details about the “plotTimeProfilesAndResiduals” task.

Example: below shows a typical mean model workflow example in which plotConfigurations and bins are set for a “plotTimeProfilesAndResiduals” task.

# Get the pkml simulation file: "MiniModel2.pkml"
simulationFile <- system.file("extdata", "MiniModel2.pkml",
  package = "ospsuite.reportingengine"
)

# Define the output object
outputA <- Output$new(
  path = "Organism|A|Concentration in container",
  displayName = "Simulated concentration of A",
  displayUnit = "µmol/ml"
)

# Define the simulation set object
setA <- SimulationSet$new(
  simulationSetName = "A",
  simulationFile = simulationFile,
  outputs = outputA
)

# Create the workflow object
workflowA <-
  MeanModelWorkflow$new(
    simulationSets = setA,
    workflowFolder = "Example-A"
  )
#> 09/09/2024 - 13:32:18
#> i Info  Reporting Engine Information:
#>  Date: 09/09/2024 - 13:32:18
#>  User Information:
#>  Computer Name: fv-az1564-176
#>  User: runneradmin
#>  Login: runneradmin
#>  System is NOT validated
#>  System versions:
#>  R version: R version 4.4.1 (2024-06-14 ucrt)
#>  OSP Suite Package version: 12.1.0
#>  OSP Reporting Engine version: 2.2
#>  tlf version: 1.5.170

# Set the workflow tasks to be run
workflowA$activateTasks(c("simulate", "plotTimeProfilesAndResiduals"))

The next chunk of code will define PlotConfiguration objects to be included in settings:

# Define plot configuration for obsVsPred
obsVsPredConfiguration <- PlotConfiguration$new(
  xlabel = "Observed values [µmol/ml]",
  ylabel = "Simulated values [µmol/ml]",
  watermark = "vignette"
)

# Define plot configuration for resHisto
resHistoConfiguration <- PlotConfiguration$new(
  xlabel = "Residuals",
  ylabel = "Count",
  watermark = "vignette"
)

# Update some fields of the configuration if necessary
resHistoConfiguration$background$plot$fill <- "lemonchiffon"
resHistoConfiguration$background$panel$fill <- "lemonchiffon"
resHistoConfiguration$background$panel$color <- "goldenrod3"
resHistoConfiguration$background$xGrid$color <- "goldenrod3"
resHistoConfiguration$background$yGrid$color <- "goldenrod3"

# Define plotConfigurations list
plotConfigurations <- list(
  resHisto = resHistoConfiguration,
  obsVsPred = obsVsPredConfiguration
)

# Define list to update task settings with plotConfigurations field and and bins field
mySettings <- list(
  plotConfigurations = plotConfigurations,
  bins = 3
)

Then, the settings field of the “plotTimeProfilesAndResiduals” task can be updated as shown below:

# Run the workflow
workflowA$plotTimeProfilesAndResiduals$settings <- mySettings

The run of the workflow with updated settings and get the resulting report as shown below.

Since the histogram across the all simulations did not get the plot configuration settings defined for individual simulations, the plots are quite different. However, the option bins is currently set among all plots of the task. Regarding observed vs simulated plots, the got the updated xlabel, ylabel and watermark as defined in the plot configuration.

# Run the workflow
workflowA$runWorkflow()
#> 09/09/2024 - 13:32:19
#> i Info  Starting run of Mean Model Workflow
#> 09/09/2024 - 13:32:19
#> i Info  Starting run of Simulation task
#> 09/09/2024 - 13:32:19
#> i Info  Splitting simulations for parallel run: 1 simulations split into 1 subsets
#> 09/09/2024 - 13:32:19
#> i Info  Starting run of subset simulations
#>   |                                                                              |                                                                      |   0%  |                                                                              |======================================================================| 100%
#> 09/09/2024 - 13:32:19
#> i Info  Simulation task completed in 0 min
#> 09/09/2024 - 13:32:19
#> i Info  Starting run of Plot Time profiles and Residuals task
#> 09/09/2024 - 13:32:19
#> i Info  Starting run of Plot Time profiles and Residuals task for A
#> 09/09/2024 - 13:32:22
#> i Info  Plot Time profiles and Residuals task completed in 0 min
#> 09/09/2024 - 13:32:22
#> i Info  Executing: pandoc --embed-resources --standalone --wrap=none --toc --from=markdown+tex_math_dollars+superscript+subscript+raw_attribute --reference-doc="C:/Users/runneradmin/AppData/Local/Temp/Rtmp8sFlI7/temp_libpath16fcb7d380f/ospsuite.reportingengine/extdata/reference.docx" --resource-path="Example-A" -t docx -o "Example-A/Report-word.docx" "Example-A/Report-word.md"
#> 
#> 09/09/2024 - 13:32:22
#> i Info  Mean Model Workflow completed in 0.1 min
Report

PK parameters plots

Selection, displayed names and units are usually managed through the ospsuite PKParameters objects or PkParametersInfo objects. The vignette “pk-parameters” provides more details about the “plotPKParameters” task.

For mean model workflows, the PK parameters are only summarized as tables. In the current version, there is no option for number of decimal numbers to set.

For population workflows, the parameters are described using plots which depends on the population workflow type. The names in the plotConfigurations list available for the “plotPKParameters” task are the following:

  • “boxplotPkParameters”: for boxplots of PK parameters performed by parallel comparison and ratio types of population workflows
  • “boxplotPkRatios”: for boxplots of PK parameters ratios performed by ratio type of population workflows
  • “vpcParameterPlot”: for VPC like plots of PK parameters vs Demography parameters performed by pediatric type of population workflows
  • “comparisonVpcPlot”: for VPC like plots of PK parameters vs Demography parameters comparing to reference population performed by pediatric type

Since pediatric workflows perform VPC like plot, aggregation of the data is performed along the demography parameters. The binning of the demography parameters can be set by the optional field bins from the task field settings. This field can be either a unique value corresponding to the number of bins or a vector defining the bin edges for all the demography parameter paths.

Besides, the final plot can either link the aggregated values or plot them as stairstep. The default behaviour is to perform a stairstep plot, but this can be tuned with the optional field stairstep (TRUE/FALSE) from the task field settings.

Below shows examples of how to set such options:

# Associate the bin edges
workflowA$plotPKParameters$settings$bins <- c(0, 1, 2, 3, 5, 10)

# Associate the number of bins
workflowA$plotPKParameters$settings$bins <- 15

# Set VPC as stair step
workflowA$plotPKParameters$settings$stairstep <- TRUE

Sensitivity plots

Settings for the sensitivity plots are SensitivityPlotSettings objects. The vignette “sensitivity-analysis” provides more details about the “plotSensitivity” task.

The input argument of the SensitivityPlotSettings settings object are then (to be created using SensitivityPlotSettings$new()):

  • totalSensitivityThreshold: cut-off used for plots of the most sensitive parameters; default is 0.9.
  • variableParameterPaths: paths that were varied in the sensitivity analysis. If supplied totalSensitivityThreshold is 1, else it uses the default or user-provided value.
  • maximalParametersPerSensitivityPlot: maximal number of parameters to display in a sensitivity plot. By default, the maximal number of parameters to display is 50.
  • plotConfiguration: PlotConfiguration object from tlf library
  • xAxisFontSize: Font size of x-axis labels for sensitivity plot, that can overwrite behavior of plotConfiguration. Default font size is 6.
  • yAxisFontSize: Font size of y-axis labels for sensitivity plot, that can overwrite behavior of plotConfiguration. Default font size is 6.
  • xLabel Label of x-axis for sensitivity plot, that can overwrite behavior of plotConfiguration. Default xLabel is “Sensitivity”.
  • yLabel Label of y-axis for sensitivity plot, that can overwrite behavior of plotConfiguration. Default yLabel is NULL.
  • colorPalette Name of a color palette to be used by ggplot2::scale_fill_brewer() for sensitivity plot. Default color palette is “Spectral”.

Since displayed sensitivity parameters can include long strings of characters shrinking the actual plot even using small yAxisFontSize, an algorithm providing line breaks is included in the task to prevent the shrinking issue as much as possible. Two fields are available in settings to tune the algorithm:

  • maxWidthPerParameter is the maximum width for the display parameters. Parameters longer than that will have line breaks. The default maximum width is 25 characters.
  • maxLinesPerParameter is the maximum number of lines for the display parameters. The default maximum number of lines is 3. For parameters longer than maxLinesPerParameter*maxWidthPerParameter, only maxLinesPerParameter setting will be respected.

Absorption

The name in the plotConfigurations list available for the “plotAbsorption” task is the following:

  • “absorptionPlot”

Mass Balance

The name in the plotConfigurations list available for the “plotMassBalance” task is the following:

  • “timeProfile”: for time profile plots of mass balance in linear and log scales
  • “cumulativeTimeProfile”: for cumulative time profile plots of mass balance in linear and log scales
  • “normalizedTimeProfile”: for normalized time profile plots (as fraction of drug mass) of mass balance in linear and log scales
  • “normalizedCumulativeTimeProfile”: for cumulative normalized time profile plots of mass balance in linear and log scales
  • “pieChart”: for pie chart of mass balance at last simulated time point

Another field of settings is selectedCompoundNames which allows to keep only a selected list of compounds to be kept in the mass balance plots.

Demography plots

The names in the plotConfigurations list available for the “plotDemography” task are the following:

  • “histogram”: for histogram of Demography parameters performed by parallel comparison and ratio types of population workflows
  • “vpcParameterPlot”: for VPC like plots of Demography parameters vs Demography parameters performed by pediatric type of population workflows
  • “comparisonVpcPlot”: for VPC like plots of Demography parameters vs Demography parameters comparing to reference population performed by pediatric type

Since pediatric workflows perform VPC like plot, aggregation of the data is performed along the demography parameters. The binning of the demography parameters can be set by the optional field bins from the task field settings. This field can be either a unique value corresponding to the number of bins or a vector defining the bin edges for all the demography parameter paths.

Besides, the final plot can either link the aggregated values or plot them as stairstep. The default behaviour is to perform a stairstep plot, but this can be tuned with the optional field stairstep (TRUE/FALSE) from the task field settings.

Similarly, parallel comparison and ratio types of population workflows perform histograms, whose binning can set using the settings field bins.