Materials for the Workshop "Hands-On Text Coding with Large Language Models for Social Scientists with R"

This repository contains the materials for the workshop "Hands-On Text Coding with Large Language Models for Social Scientists with R".

Requirements and setup

Below I describe the software requirements ...

• R
• RStudio
• required R packages
• ollama

... and necessary setup steps.

• downloading the 8 billion LlaMa 3 model with rollama
• creating an OpenAI (Plus) account
• getting an OpenAI API key

If you run into difficulties, email me at hauke.licht [at] wiso [dot] uni-koeln [dot] de or post an issue.

R

One main prerequisite for the workshop is that you have programmed in R before. Hence, I am assuming that you have a working R installation on the computer you will be using during the workshop.

• Please make sure that your R version is 4.2.0 or higher: type R.version in your R console or R --version in your Terminal app (MacOS users) or Command Prompt (Windows users)
• If needed, download and install R as described here: https://posit.co/download/rstudio-desktop/

RStudio

We will be using RStudio as a code editor in the workshop. Please install RStudio before the workshop if neede as described here https://posit.co/download/rstudio-desktop/

Required R packages

Next, you need to install install all required packages. I recommend you use renv for this:

1. Open this folder as a project in RStudio (see here how to)
2. open the R script setup/setup_renv.R
3. Run the complete script.

Alternative: If you don't want to use renv for installing and managing required packages, you can mannually install the packages (in the versions) listed in file setup/requirements.txt

ollama

We will be using ollama to interact with open-source LLMs. Please install ollama from here: https://ollama.com/download

Note — You will need macOS 11 Big Sur (or later) or Windows 10 (or later).

Using ollama in R

Run the code in notebook test_ollama.qmd to verify that you can access API key is accessible in RStudio and your API access works.

Open AI account and API access

1. Create an OpenAI developer account

Go to OpenAI and create an account.

2. Link a payment method and book some credit

1. go to platform.openai.com/account/billing/overview
2. link a valid credit card
3. click "Add to credit balance" and load a credit of, e.g., U.S.$ 10

3. Get your API key

Go to the API keys page.

Create a new API key by clicking on the "Create new secret key" button.

Fill your information in the form show below ...

... and click on the "Create secrete " button.

Note — This will be the only time you will be able to see your API key. Make sure you take the next step described below to store it in a safe place.

4. Make your API key accessible in RStudio

Create a file called .Renviron in the root of your project folder. (It's important that the file name starts with a dot!) Open it in a text editor

Alternative — Directly create and open the .Renviron file by running usethis::edit_r_environ(scope = "project") in your R console in RStudio (see https://github.com/irudnyts/openai?tab=readme-ov-file#authentication).

Now,

1. add OPENAI_API_KEY= in the first line of the file,
2. copy your API key from your browser window,
3. paste the key in your .Renviron file behind the =
4. save the file and close it

5. Verify that your API key is accessible in python

Run the code in notebook test_openai_key.qmd to verify that your API key is accessible in RStudio and your API access works.

Questions and issues

If you have a question or encounter any issues, email me at hauke.licht [at] wiso [dot] uni-koeln [dot] de or post an issue.