An Interactive Specification Curve — shiny_rdf_spec

A shiny based web app that allows you to explore your researcher degrees of freedom's specification curve interactively.

Usage

shiny_rdf_spec_curve(
  ests,
  spec_curve_parms,
  spec_curve_selected = NULL,
  design = NULL,
  rel_dir = NULL,
  start_input = NULL,
  libs = NULL,
  add_files = NULL,
  regression_cutoff = 5,
  model_render_func = NULL,
  default_choices = NULL,
  restore_button = FALSE,
  title = "A Shiny Specification Curve",
  abstract = NULL,
  choice_labels = NULL,
  with_spinner = FALSE,
  spinner_options = list()
)

Arguments

ests: The data frame provided by exhaust_design.
spec_curve_parms: A list containing additional parameters that will be passed on to plot_rdf_spec_curve. You can also provide a named list of lists with each list containing parameters for a specification curve. In this case, the shiny app will include an additional Select Input element where the users can select which specification curve to display.
spec_curve_selected: If you provide a named list of specification curves in spec_curve_parms, you can provide here the name of the default curve to plot first.
design: if not NULL it takes the design that was used to generate the estimates. In this case, you either need to have all required design elements in your current environment or you need to specify the rel_dir parameter pointing to the code files below. In addition, you need to set start_input. The shiny app will then display regression results when you select choices that generate less than regression_cutoff estimates.
rel_dir: The path to the code directory where the design functions are located. Only needed when the functions are loaded to your current environment See above.
start_input: The parameters that you pass to the first design step. See above.
libs: A vector containing additional packages that need to be attached to run the design. NOTE: This will modify the shiny app code to include literal library() calls so that shinyapps.io includes the libraries on deployment.
add_files: A character vector containing relative paths to files and directories that you want to bundle with the shiny app. The files will be copied to the temporary directory that hosts the shiny app and directories will be copied recursively.
regression_cutoff: If your choices generate less or equal estimates, the display will switch to normal regression table output (needs parameters above to be not NULL).
model_render_func: A function to create the regression table, taking a list of the models as parameter. The function is evaluated within the shiny app environment. By default (NULL), the regressions are rendered by calling the internal function renderModels() that then calls modelsummary to create the HTML output. If you need to prep the model data for preparation, you can provide a function here that calls renderModels() after prepping the data. Alternatively, you can provide a function that generates the HTML output directly.
default_choices: A list containing choices that you want the app to start with. If NULL, it will start with all choices included.
restore_button: Set to TRUE when you want to have a restore button in the app (defaults to FALSE).
title: The title of the shiny app.
abstract: Text that will be displayed by the app. Wrapped into HTML() so that you can use HTML code.
choice_labels: Character vector containing the labels that will be used to label the select list input controls in the shiny app. If NULL, the select list input controls are labeled based on the choice column names from the ests data frame.
with_spinner: Do you want to include a spinner (useful when displays take some time to render). Defaults to FALSE. See withSpinner for detail.
spinner_options: A list containing parameters that you want to to pass to withSpinner.

Examples

if (FALSE) {
  print("Sorry. No examples yet.")
}