blockr.ai

Overview

blockr.ai provides AI-powered blocks for the blockr.core framework. It enables integration with various AI services and models to enhance data analysis workflows through natural language interactions.

Features

• Natural language powered plot creation through new_llm_plot_block()
• AI-assisted data transformations through new_llm_transform_block()
• Integration with OpenAI’s GPT models
• Interactive blocks that can be composed with other blockr components

Installation

You can install the development version of blockr.ai from GitHub with:

# install.packages("remotes")
remotes::install_github("cynkra/blockr.ai")

Usage

blockr.ai extends the blockr.core framework with AI capabilities.

Examples

library(blockr.core)
library(blockr.ai)

serve(
  new_llm_plot_block(),
  data = list(`1` = iris)
)

serve(
  new_llm_transform_block(),
  data = list(`1` = iris)
)

serve(
  new_board(
    blocks = blocks(
      a = new_dataset_block("iris"),
      b = new_llm_plot_block(
        "Createa a scatterplot of Sepal.Length against Sepal.Width.",
        paste(
          "ggplot2::ggplot(dataset_1) +",
          "  ggplot2::geom_point(",
          "    ggplot2::aes(x = Sepal.Width, y = Sepal.Length)",
          "  )",
          sep = "\n"
        )
      )
    ),
    links = links("a", "b")
  )
)

License

GPL (>= 3)