FIRST-JULIA-NN

1. Get the full dataset inside the data directory. (Data_Julia subdir and Completed_SAO_V3.hdf5 file)
2. Install requirements

pip install -r requirements.txt

3. Create partitions

python main.py root=(pwd)

4. Generate folds and run study

python main.py root=(cwd) action=run_study

To train using CPU:

python main.py root=(cwd) action=run_study training.device=cpu

Other instructions

1. Compile C++ (Arrow/Plasma/Parquet)

clang++ src/test.cc --std=c++17 -larrow -lplasma -lparquet