# FinalProjectAntarctis
# Code repository

This is the Hugging Face for Adian Khair Oellegaard, Jens Gorm Rytter, Markus Bligaard Winther, Phillip Nordfang, Silas Larsen. Here, we can structure and connect our work.

## Description
This repository is for the Applied Machine Learning 2025- Antarctic ice sheet ice thickness project. A parquet datafile of different icesheet thickness measurements along with other features is made in the create_bedmap_dataset.ipynb file. Our ML models should be compared to the glaciological model BedMachine v3.7 which thickness predictions are given in the datafile in the column: 'ith_bm'.

(Note: The datafiles are uploaded to a separate repository, called finalprojectdata)

(Note: The placement of our datafiles and code files might be different and so you might have to change paths in the code)
## Getting Started

### If you want to to hook up the repository directly to a folder:

Step 0:
Make sure you have a Hugging Face account where finalprojectantarctis is shared with you.

Step 1:
Go to the folder where you want the code stored.

Step 2: Open Git Bash in this folder and write:

pip install huggingface_hub

Step 3: Generate a key for your account
On Hugging Face go to settings->Access Tokens: Then give the key a name and select write role.

Step 4: Login to Huggin Face user through git bash:
huggingface-cli login

Step 5: Clone repository
git clone https://huggingface.co/finalprojectantarctis/Code


Upload/Update code file:
git add <codefilename>
git commit -m "quick description of update"
git push

Download code files:
git pull

(Ask ChatGPT (or someone in the group) if you get stuck)


### Alternatively use browser
You can also download and add files to the repository through a browser with the link: https://huggingface.co/finalprojectantarctis/Code/edit/main/README.md
### Dependencies

* Prerequisites for running the code should be written in requirements.txt

## Authors

Adian Khair Oellegaard
Jens Gorm Rytter
Markus Bligaard Winther
Phillip Nordfang
Silas Larsen

## Acknowledgments

Inspiration, code snippets, etc.