Title: Data Visualization Utilities for 'pyDarwin' Machine Learning Pharmacometric Model Development
Version: 2.0.1
Description: Utilize the 'shiny' interface for visualizing results from a 'pyDarwin' (https://certara.github.io/pyDarwin/) machine learning pharmacometric model search. It generates Goodness-of-Fit plots and summary tables for selected models, allowing users to customize diagnostic outputs within the interface. The underlying R code for generating plots and tables can be extracted for use outside the interactive session. Model diagnostics can also be incorporated into an R Markdown document and rendered in various output formats.
URL: https://certara.github.io/R-DarwinReporter/
Depends: R (≥ 4.1.0)
License: LGPL-3
Encoding: UTF-8
RoxygenNote: 7.3.2
Suggests: knitr, rmarkdown, data.table, readr, testthat (≥ 3.0.0)
Imports: DT, colourpicker, shinyAce, shinymeta, utils, ggplot2, xpose, Certara.Xpose.NLME, dplyr, jsonlite, tidyr, flextable, shinyjqui, grDevices, plotly, scales, shiny (≥ 1.7.0), shinyjs, shinyWidgets, bslib (≥ 0.7.0), shinyTree (≥ 0.3.1), sortable
Config/testthat/edition: 3
NeedsCompilation: no
Packaged: 2025-03-06 16:41:17 UTC; jcraig
Author: James Craig [aut, cre], Michael Tomashevskiy [aut], Mike Talley [aut], Certara USA, Inc [cph, fnd]
Maintainer: James Craig <james.craig@certara.com>
Repository: CRAN
Date/Publication: 2025-03-07 16:40:13 UTC

Generate and Report Model Diagnostics from NLME or NONMEM runs

Description

Shiny application to generate, customize, and report diagnostic plots and tables from NLME or NONMEM output files. Create an Rmarkdown file of tagged model diagnostics and render into submission ready report.

Usage

darwinReportUI(darwin_data, tagged = NULL, settings = NULL, ...)

Arguments

darwin_data

Object of class darwin_data. Note, key_models xpose_data must be available.

tagged

List of tagged objects returned from previous tagged <- darwinReportUI() session.

settings

List of settings (e.g., settings.Rds) returned from previous Shiny session.

...

Additional arguments for Pirana integration.

Value

If interactive(), returns a list of tagged diagnostics from the Shiny application, otherwise returns TRUE.

Examples

if (interactive()) {
ddb <- darwin_data("./darwin_search_09") |>
   import_key_models("./darwin_search_09/key_models")

darwinReportUI(ddb)
}

Initialize darwin data structure.

Description

Initialize darwin data structure.

Usage

darwin_data(
  project_dir,
  working_dir = NULL,
  output_dir = NULL,
  key_models_dir = NULL,
  ...
)

Arguments

project_dir

Directory containing input files for pyDarwin (e.g., options.json).

working_dir

Directory containing misc results folders generated from a pyDarwin search. This is the default location of the key_models, output, and temp folders.

output_dir

Directory containing output files such as "results.csv" and final control files. Default location is inside working_dir/output.

key_models_dir

Directory of the key_models folder. Default location is inside working_dir/key_models. Note, key models are not imported if argument is NULL, explicitly specify key_models_dir to import files for darwinReportUI.

...

Additional args.

Details

If working_dir and output_dir are sub directories of project_dir, these arguments can be ignored. The key_models_dir is not required to initialize the darwin_data object. If specified, however, key models data will be imported which may take time given the number of key models and size of output tables. See import_key_models.

Value

Object of class darwin_data.

Plot minimum fitness by iteration with penalty composition.

Description

Plot minimum fitness by iteration with penalty composition.

Usage

fitness_penalties_vs_iteration(
  darwin_data,
  group_penalties = TRUE,
  scale_ofv = TRUE,
  ...
)

Arguments

darwin_data

Object of class darwin_data.

group_penalties

Logical; defaults to TRUE.

scale_ofv

Set to TRUE to rescale OFV axis limit. Used to better observe penalty effects.

...

Additional arguments.

Value

Object of class ggplot.

Plot best fitness by iteration.

Description

Plot best fitness by iteration.

Usage

fitness_vs_iteration(darwin_data, ...)

Arguments

darwin_data

Object of class darwin_data.

...

Additional arguments.

Value

Object of class ggplot.

Get eps shrinkage values xpose_data object

Description

This function returns eps shrinkage values from xpose_data object as a data.frame.

Usage

get_eps_shk(xpdb)

Arguments

xpdb

Object of class xpose_data.

Value

Returns an object of class data.frame.

Get eta shrinkage values from xpose_data object

Description

This function returns eta shrinkage values from xpose_data object as a data.frame.

Usage

get_eta_shk(xpdb)

Arguments

xpdb

Object of class xpose_data.

Value

Returns an object of class data.frame.

Imports files from key model output folders

Description

Use to create xpose data object from files in NLME or NONMEM key model output folders.

Usage

import_key_models(darwin_data, dir, ...)

Arguments

darwin_data

Object of class darwin_data.

dir

File path to key models directory.

...

Additional args.

Value

Object of class darwin_data.

Examples

if (interactive()) {
ddb <- darwin_data(project_dir = "./darwin_search_09") |>
   import_key_models(dir = "./darwin_search_09/key_models")
}

Summarise fitness by iteration

Description

Summarise minimum, cumulative minimum, and mean fitness values by pyDarwin search iteration/generation.

Usage

summarise_fitness_by_iteration(darwin_data)

Arguments

darwin_data

Object of class darwin_data.

Value

data.frame with columns iteration, min_fitness, mean_fitness, and min_cum_fitness

Summarize minimum fitness and penalty values by iteration

Description

Summarise minimum fitness, ofv, and penalty values used in calculation of overall fitness values by pyDarwin search iteration/generation.

Usage

summarise_fitness_penalties_by_iteration(darwin_data, group_penalties = FALSE)

Arguments

darwin_data

Object of class darwin_data.

group_penalties

Logical. Set to TRUE to group penalties.

Value

data.frame of columns "iteration", "fitness", "ofv" and corresponding penalty columns.

Summarise overall table by key models

Description

Generate a summary data.frame by key models, which includes columns such as condition number, number of parameters, -2LL, AIC, BIC, fitness, and penalty values.

Usage

summarise_overall_by_key_models(darwin_data)

Arguments

darwin_data

Object of class darwin_data.

Value

data.frame

A ggplot2 theme for Certara.

Description

A ggplot2 theme for Certara.

Usage

theme_certara(
  base_size = 11,
  base_family = "",
  base_line_size = base_size/22,
  base_rect_size = base_size/22,
  grid = c("none", "horizontal", "both"),
  ...
)

Arguments

base_size

base font size, given in pts.

base_family

base font family

base_line_size

base size for line elements

base_rect_size

base size for rect elements

grid

Which grid lines should appear? Horizontal only, both horizontal and vertical, or none (default). continuous_scale().

...

Additional args

Details

There are 3 variants of the theme: no grid theme_certara(), full grid theme_certara(grid = "both"), and horizontal grid lines only theme_certara(grid = "horizontal").

Value

An object of class theme().