Skip to contents

Setup Python Environment

Source: R/python_setup.R
setup_python_env.Rd

Intelligently sets up Python virtual environment with required packages. Detects existing Python installations and guides users if Python is missing.

Usage

setup_python_env(envname = "textanalysisr-env", force = FALSE)

Arguments

envname

Character string name for the virtual environment (default: "textanalysisr-env")

force

Logical, whether to recreate environment if it exists (default: FALSE)

Value

Invisible TRUE if successful, stops with error message if failed

Details

This function:

  • Automatically detects if Python is already installed

  • Offers to install Miniconda if no Python found

  • Creates an isolated virtual environment (does NOT modify system Python)

  • Installs ONLY 6 core packages (minimal installation):

    • langchain-core (core LangChain functionality)

    • langchain-ollama (Ollama integration)

    • langgraph (workflow graphs)

    • langgraph-checkpoint (workflow state management)

    • ollama (Ollama client)

    • pdfplumber (PDF table extraction)

  • Dependencies installed automatically by pip

  • Avoids heavy packages (no marker-pdf, nougat-ocr, torch)

The virtual environment approach means:

  • No conflicts with other Python projects

  • Easy to remove (just delete the environment)

  • System Python remains untouched

  • Much smaller download (~100MB vs 5GB+)

After setup, restart R session to activate enhanced features.

Examples

if (FALSE) { # \dontrun{
# First time setup (auto-detects Python)
setup_python_env()

# Recreate environment
setup_python_env(force = TRUE)
} # }