Codyan

Large Language Model Qualitative Coder Application.

Use this to apply pre-made codes with descriptions to pieces of text, for example labelling questionnaire responses.

Loosely based on the methods defined by Tai et al. and Xiao et al.

Usage

1. Make a copy of /input_sample directory named /input
2. Update the example files inside the /input directory with your own data
   1. Enter your codebook in codes.csv
   2. Enter the data you want to code in responses.csv
   3. Enter your reference, human-coded lines, if you have any, in reference.csv - this is for intercoder reliability calculations
3. Install dependencies with pip install -r requirements.txt
4. Run the Codyan with python app.py
5. Navigate to http://127.0.0.1:8000
6. Obtain and enter your OpenAI API key
7. Follow the on-screen instructions to load your input data, your codeset, perform coding and analyse your results
8. Find your results in the /output folder and on-screen. The used prompts are also saved in /output/prompts.csv

With any questions, feel free to contact me at [email protected]

Docker

If you prefer to run Codyan as a docker container, you can build the image and run it as follows:

1. Build image with docker build --tag codyan .
2. Run image with docker run -d -p 8000:5000 codyan where 8000 is the port you will access the application at