DataPrism

Suggested Python: 3.9.13

Reproducibility Steps

Step 0: Download and Install git lfs

Follow the instructions here: https://git-lfs.github.com/

Once downloaded and installed, set up Git LFS for your user account by running:

git lfs install

You only need to run this once per user account.

Step 1: Download this repo and cd to it

git clone https://github.com/sainyam/DataPrism
cd DataPrism
git lfs pull

Step 2: Download datasets.zip

https://drive.google.com/file/d/1syQhwIRwdWJBqqT0mJQWfN9LGsZOylnV/view?usp=sharing
unzip datasets.zip

Step 3: Install dependencies

sudo apt-get install -y virtualenv
virtualenv venv
source ./venv/bin/activate 

pip install -r requirements.txt

In case you want to use a custom location for python, please use virtualenv -p <PYTHON_PATH> venv instead

Step 3: Generate all plots

bash reproduce.sh

One of the baselines (Anchors) take several hours to run for all experiments. To avoid running only this baselline, please run any script with flag -a off. Example:

bash reproduce.sh -a off

Plots will be present in freshRuns/ folder