Apps

• Packages

  • {statlingua} - Designed to help bridge the gap between complex statistical outputs and clear, human-readable explanations. By leveraging the power of Large Language Models (LLMs), statlingua helps you effortlessly translate the dense jargon of statistical models — coefficients, p-values, model fit indices, and more — into straightforward, context-aware natural language.
  • {querychat} for both R and Python
    • Python Shiny template, Joe Cheng’s app, Machlis article
    • You provide metadata about your data (i.e. data dictionary stuff like column types, data description)
    • Send that and the natural language question and the LLM sends back the SQL query which gets executed locally
    • Typically used in apps and dashboards (shiny) to filter data that’s displayed

• Resources

  • Dancho Agent Examples using {langchain}

    • Automated Data Cleaning Agent (Video, Code)
    • Automated Data Wrangling Agent (Video, Code)
    • Automated Feature Engineering Agent (Video, Code)
    • Automated Data Visualization Agent (Video, Code)
    • Automated SQL Agent and App (Agent Video, Agent Code, App Video, App Code)
    • Automated H2O Machine Learning Agent (Video, Code)
    • Automated MLFlow Agent (Video, Code)

• Apps

  • R and R Studio Tutor - GPT Studio for R or R Studio students and users
    • Ask a question about R or R Studio or upload code that needs review or debugging
  • Posit Shiny AI Assistant - Sends prompts to Anthropic Claude 3.5 Sonnet (i.e. not private)
    • Intro
    • Run locally (github)
    • Hosted App
  • {chores} - Add-in that helps you automate tedious tasks. Store a prompt in a markdown file that performs a task. Bind a word or phrase to that task. Then open that add-in ui, type the word/phrase, and the llm takes care of the rest. Preloaded chores:
    • “cli”: Convert to cli
    • “testthat”: Convert to testthat 3
    • “roxygen”: Document functions with roxygen

• Example: Simple {shinychat} Chat Interface (source)

  library(shiny)
  library(shinychat)
  
  ui <- bslib::page_fluid(chat_ui("chat"))
  server <- function(input, output, session) {
    chat <- ellmer::chat_ollama(model = "gemma2")
  
    observeEvent(input$chat_user_input, {
      stream <- chat$stream_async(input$chat_user_input)
      chat_append("chat", stream)
    })
  }
  shinyApp(ui, server)