Installation

1. Clone the repository

   git clone https://github.com/sustainable-processes/deep_gamma.git

2. Get the data

   TO-DO

3. Initialize chemprop submodule

   git submodule init  
   git submodule update

Data generation

See README indeep_gamma/data/cosmo

Preprocessing

python deep_gamma/graphs/data_preprocessing.py 

To process polynomial data

python deep_gamma/ops/polynomial.py fit data/02_intermediate/cosmo_data.pq data/03_primary/polynomial
python deep_gamma/ops/polynomial.py split data/03_primary/polynomial data/05_model_input/cosmo data/05_model_input/cosmo

Machine Learning

COSMO

python deep_gamma/ops/chem.py --data_dir data/ --experiment_name cosmo_base

python deep_gamma/ops/chem.py --data_dir data/ --experiment_name cosmo_base_pretrained --wandb_checkpoint_frzn_run=2532qdqg

COSMO polynomial

python deep_gamma/ops/chem.py --data_dir data/ --experiment_name cosmo_polynomial  --dataset cosmo-polynomial

Combisolv

python deep_gamma/ops/chem.py --data_dir data/ --experiment_name cosmo_base --artifact_name cosmo_base --combisolv

Fingerprints

python deep_gamma/ops/fingerprint.py --wandb_checkpoint_run 32vlsf8l --results_path data/ --dataset aspen

Then rename to data/07_model_output/aspen to aspen_base_fingerints

python deep_gamma/ops/fingerprint.py --wandb_checkpoint_run 3s8jnyvi --results_path data/ --dataset aspen

Then rename to data/07_model_output/aspen to aspen_pretrained_fingerints

Eval

python deep_gamma/ops/eval.py --results_path=data/ --drop_na