placeholders

You may wish create your report template with placeholders. For example if you want create a generic template for population PK analysis, it may be useful to label the figures generically. You can use this section to define placeholders that will be applied when the report is created. For example consider the yaml code here:

placeholders:
  CMPD: Compound Name
  CUNITS: Conc Units
  OBJ: sprintf("%3g", fit$objf)

When you create the report you can label your y-axis generically such as:

"===CMPD=== (===CUNITS===)"

But when the report is built, it will become:

"Compound Name (Conc Units)"

Notice that the OBJ placeholder contains R Code. By default the reporting process will try to evaluate the placeholders as R Code. If that fails it will revert back to the string supplied. If you run into problems you can quote your placeholder value with both double and single quotes. For example:

  ISSTR:"'ls()'"
  ISCMD: ls()

The ISSTR field will just result in the string "ls()" while the ISCMD field will return a list of the objects in the environment where the string is evaluated. It is also possible to overwrite the contents of this file at runtime by providing a placeholders (placeholder) argument to report_fit().

Note: Fields that support placeholders will be indicated by ^[PH] and fields with evaluable code will be indicated by ^[EV].

parameters

For certain reporting elements it may be better to present parameters differently than the parameter names being used in the model. In nlmixr you can use a comment after the parameter definition to specify alternate text. The function gen_pest_table() will use this alternate text if it is present. In the parameters section you have some more flexability in how parameters are reported. You can define any parameters you want here whether they are present in the model or not. They will only be used if they are present. For tables in Word or PowerPoint you can specify the names using markdown (md), otherwise you can specify how they will be handled in text environments (txt). For example the intrinsic clearance may be called lCLint in the model. And you can specify it as:

parameters:
  lCLint:    
    md:  "CL~int~ (L/hr)"    
    txt: "CLint (L/hr)"    

It is also possible to overwrite the contents of this file at runtime by providing a parameters argument to report_fit().

Note: If you specify a parameter name here it will overwrite any alternate text specified in the model ini().

covariates

For figure and table generation it can be useful to have your covariates defined. You can define them here as either categorical (cat) or continuous (cont). For example to define the categorical covariates as SEX and ROUTE and the continuous covariate as WT you would do the following:

covariates:
  cat: ["SEX", "ROUTE"]
  cont: ["WT"]

If you don’t need any covariates you can set them to NULL. If you want to overwrite these at runtime you can specify different values when calling report_fit(). These are available in the figure and table runtime environments as the objects cat_covars for categorical covariates and cont_covars for continuous covariates.

options

This section allows you to control general report options. Each element here can be either an explicit value. If identified as evaluable ^[EV] it will first attempt to evaluate it as R code. If that fails the explicit value will be used. The following options are supported with the expected data type in parentheses:

• output_dir^[PH,EV] - The location where figures will be generated (character)

• resolution^[EV] - Resolution of figures being generated (numeric)

• figenv_preamble and tabenv_preamble - User-defined code evaluated before figures and tables are built (specify libraries needed, source files containing user defined code, etc.)

• fig_stamp^[PH] - If this is set to a string it will stamp the contents of that string at the lower left portion of the figure after writing. This applies only to figures where the p_res object is either a vanilla ggplot object, a paginated ggforce object, or the result of ggarrage from the ggpubr package. Along with the regular placeholders, this option can also contain the ===FILE=== placeholder. This will be replaced with the full path to the figure file that is generated.

Building figures & tables

Figures and tables are defined in terms of R code that will be evaluated. There are several objects available in the environment that can be used when creating these report elements. The following objects will be defined:

• fid - current figure ID (figure environment only)
• tid - current table ID (table environment only)
• fit - nlmixr2 fit object
• cat_covars - character object containing the categorical covariates
• cont_covars - character object containing the continuous covariates
• height - figure height in inches (figure environment only)
• obnd - onbrand object of the report document
• output_dir - directory where figures are stored
• resolution - resolution of the figure (figure environment only)
• rptdetails - result of reading in the report yaml file
• rpttype - either “Word” or “PowerPoint”
• width - figure width in inches (figure environment only)
• xpdb - the default yaml file contains code in the preamble to create this object by running xpose_data_nlmixr(fit) (figure environment only)

figures

The figures section consists of figure IDs. These IDs are used again when defining the Word and PowerPoint report contents. Consider this example:

figures:
  dv_vs_pred:
    orientation:    "portrait"
    caption:        "dv_vs_pred caption"
    caption_format: "text"
    title:          "dv_vs_pred title"
    cmd: |-
      p_res <- dv_vs_pred(xpdb, caption=NULL, title=NULL) +
      xlab("Observed ===CMPD=== Concentrations (===CUNITS===)") +
      ylab("Population Predicted ===CMPD=== Concentrations (===CUNITS===)")

It creates the figure with the ID dv_vs_pred. It has the following elements:

• orientation - Orientation when used in a Word report, it can be either “portrait” or “landscape”.
• caption^[PH] - Figure caption in Word reports.
• caption_format - This element is optional and can be either "text"(default plain text) or "md" (Markdown).
• title^[PH] - Slide title when the figure is used in PowerPoint.
• cmd^[PH] - Command to generate the figure.

Note: When the cmd is evaluated it needs to create a variable called p_res. This can be either a ggplot object, a paginated object from ggforce, the result of ggarrage from the ggpubr package, or a vector of file names containing image files created by the code in cmd. If you want to conditionally skip the reporting of a figure, you can instead set the value of p_res to NA. This will generate a message and prevent inclusion of a build error in the final document. You should also include any libraries you want in the relevant preamble section to ensure the functions are available to you.

tables

Tables behave similarly to figures. In the tables section you create a table ID with the following options specified.

• orientation - Orientation when used in a Word report, it can be either “portrait” or “landscape”.
• caption^[PH] - Table caption in Word reports.
• caption_format - This element is optional and can be either "text"(default plain text) or "md" (Markdown).
• title^[PH] - Slide title when the table is used in PowerPoint.
• cmd^[PH] - Command to generate the table.

This example will create a table of the parameter estimates.

tables:
  pest_table: 
    orientation:    "portrait"  
    caption:        "Parameter Estimates"
    caption_format: "Parameter Estimates"
    title:          "Parameter Estimates"   
    cmd: |- 
      t_res <- gen_pest_table(
      fit        = fit, 
      obnd       = obnd,
      rptdetails = rptdetails)

Note: When the cmd is evaluated it needs to create a variable called t_res. This a list which can contain the following elements:

• df - List of data frames containing the tabular information.
• ft - List of flextables containing the tabular information

If you only supply the data frames list element, they will be converted into flextables internally. Flextables will span multiple pages automatically. If you prefer to have a caption on each page you can split the tables manually into multiple data frames/flextables. The reporting functionality is written to support table spanning multiple pages. If you define a separate table for each page you need to add those as sequential list elements to the df and ft elements. For example:

t_res$df = list()
t_res$df[[1]] = data.frame()
t_res$df[[2]] = data.frame()
etc

or

t_res$ft = list()
t_res$ft[[1]] = flextable::flextable() 
t_res$ft[[2]] = flextable::flextable() 
etc

If you only have a single table or if you want it to span multiple pages automatically you simply specify the first element listed above (e.g. t_res$df[[1]]).

If you want to conditionally skip the generation of a table, simply set t_res to NA. Like with figures above, this will signal to the reporting functions that you do not want to generate this table for the current fit object. It will also prevent the generation of a build error.

pptx

This section defines general aspects of PowerPoint reports and also the contents of the report relating to the fitting results. Both the figures and tables sections contain the onbrand slide master names ("content_text") to be used to hold these objects. The slides will be created with a title specified in figure and table IDs. The placeholder name for the title ("title") must be specified as well as the placeholder name for the content ("content_body"). Figures for PowerPoint reports must be generated in the dimensions of the placeholder, so it is important that the width and height elements are specified in inches.

pptx:
  figures:
    master:   
      name:       "content_text"      
      title_ph:   "title"   
      content_ph: "content_body"
    width:  9.5
    height: 5.0
  tables:  
    master:   
      name:       "content_text"      
      title_ph:   "title"   
      content_ph: "content_body"

The actual content is specified in the content section. This is a list of keywords followed by a value. The table keyword should be followed by a table ID defined in the tables section. Similarly the figure keyworld should be followed by a figure ID defined in the figures section:

  content:  
    - table: pest_table
    - figure: ind_plots
    - figure: dv_vs_pred
    - figure: dv_vs_ipred
    - figure: res_vs_pred
    - figure: res_vs_idv
    - figure: prm_vs_iteration
    - figure: absval_res_vs_pred

docx

This section starts with general information about word documents. The figures section defines the height and width of figures for both landscape and portrait orientations. Again the units here are inches.

docx:
  figures:
    landscape:
      width:  8.0
      height: 4.2
    portrait:
      width:  6.5
      height: 6.0

Just like with PowerPoint the content is defined in the content section. This can be defined as a figure or a table the same as in PowerPoint. Content can also include text elements as well. Just specify the onbrand style to control general formatting.

  content:
    - text: 
        text:  "Figures"
        style: Heading_1
    - figure: dv_vs_pred
    - figure: ind_plots
    - figure: dv_vs_ipred
    - figure: res_vs_pred
    - figure: res_vs_idv
    - figure: prm_vs_iteration
    - figure: absval_res_vs_pred
    - text: 
        text:  "Tables"
        style: Heading_1
    - table: pest_table

PowerPoint

To create an onbrand PowerPoint template for your organization you will need the following master slide names define with the corresponding elements.

Word

Similarly, to create an onbrand Word template you will need to the following onbrand styles defined.