rk.shiny.plugins

RKWard Shiny Plugins Package and Generator

Synopsis

This repository contains a single R script, make_rk_shiny.R, designed to programmatically generate a complete RKWard plugin package using the rkwarddev R package.

The resulting R package, rk.shiny.plugins, bundles several interactive data visualization and analysis tools (Shiny apps and HTML widgets) into a single, easy-to-install RKWard extension. The entire development process is encapsulated within the R script, making it reproducible, easy to maintain, and simple to extend with new plugins.

Features

• Programmatic Generation: The entire plugin structure, including all .xml, .js, and help files, is generated from a single R script. There is no need to manually edit XML or other plugin files.
• Single, Unified Package: All interactive plugins are bundled into one package for easy distribution and installation.
• Categorized Menu Structure: Plugins are organized logically under a top-level "Shiny" menu (Visualization, Exploration, Statistics, Psychometrics).
• Modern & CRAN-Compliant: Completely rebuilt to remove archived legacy dependencies (like rpivotTable), replacing them with modern, stable alternatives.
• Multi-language Support (i18n): Fully translated into Spanish, French, German, and Brazilian Portuguese.

Plugins Included

This package currently generates the following plugins, accessible from the Shiny menu in RKWard:

📊 Visualization

• Interactive Pivot Table: A powerful drag-and-drop pivot table. It launches automatically in your external web browser to ensure maximum stability and prevent websocket timeouts. (Depends on shinypivottabler and shiny).
• Professional Pivot Table: Generates highly customizable, publication-ready static HTML pivot tables inside the RKWard output. Includes native support for Weighted Sums and Means. (Depends on pivottabler).
• Visual Explorer: A modern, "Tableau-style" drag-and-drop data visualization interface perfect for fast visual analytics. (Depends on GWalkR).
• ggplot GUI: The classic "point-and-click" interface for building ggplot2 graphics. (Depends on ggplotgui).
• Esquisse Plot Builder: A drag-and-drop builder for ggplot2 to easily create charts and export the underlying code. (Depends on esquisse).

🔍 Exploration

• Automated EDA Report: Automatically generates a complete HTML data profiling report (missing values, correlations, histograms) in one click. (Depends on DataExplorer).
• Quick EDA: A comprehensive Shiny gadget for fast exploratory data analysis and filtering. (Depends on ggquickeda).

📈 Statistics

• Factoshiny: Interactive multivariate analysis (PCA, CA, MCA) with dynamic graph customization and clustering. (Depends on Factoshiny).
• Shinystan Diagnostics: Interactive visual diagnostics for Bayesian models (MCMC chains, posterior distributions, R-hat). (Depends on shinystan).

🧠 Psychometrics

• Shiny Item Analysis: A comprehensive suite for Classical Test Theory (CTT) and Item Response Theory (IRT) analysis. (Depends on ShinyItemAnalysis).

How to Use

Follow these steps to generate, compile, and install the complete plugin package.

Prerequisites

Ensure you have R, RKWard, and the necessary R packages installed. You can install all required R packages by running this command in your R console:

install.packages(c(
  "devtools", "rkwarddev", "shiny",
  "shinypivottabler", "pivottabler", "GWalkR", 
  "ggplotgui", "esquisse", "ggquickeda", "DataExplorer", 
  "Factoshiny", "shinystan", "ShinyItemAnalysis"
))

Troubleshooting: Errors installing devtools or missing binary dependencies (Windows)

If you encounter errors mentioning "non-zero exit status", "namespace is already loaded", or requirements for compilation (compiling from source) when installing packages, it is likely because the R version bundled with RKWard is older than the current CRAN standard.

Workaround: Until a new, more recent version of R (current bundled version is usually 4.3.3) is packaged into the RKWard executable, these issues will persist. To fix this:

1. Download and install the latest version of R (e.g., 4.5.x or newer) fromCRAN.
2. Open RKWard and go to the Settings (or Preferences) menu.
3. Run the "Installation Checker".
4. Point RKWard to the newly installed R version.

This "two-step" setup (similar to how RStudio operates) ensures you have access to the latest pre-compiled binaries, avoiding the need for RTools and manual compilation.

Installation

If you have installed RKWard >= 0.7.4, you can use the install functionality directly from git. As shown here: Shiny apps on RKWard

You just need to add:

• User name: "AlfCano"
• Repository: "rk.shiny.plugins"

Or, just run the next command in your R console:

local({
  ## Prepare
  require(devtools)
  ## Install
  install_github(
    repo="AlfCano/rk.shiny.plugins"
  )
  ## Print Result
  rk.header("Results of installing from git")
})

Generator

To run the generator script yourself (if modifying the plugins), the rkwarddev package is needed.

install.packages("rkwarddev")

Author

• Alfonso Cano (alfonso.cano@correo.buap.mx)
• Assisted by Gemini, a large language model from Google.