environment.yml

name: applied-ai
channels:
  - defaults
  - conda-forge
  - pytorch
dependencies:
  # General dependencies
  - python==3.8.*
  - scipy==1.6.*
  - pandas==1.*
  - scikit-learn==1.0.*
  - matplotlib==3.*
  - tensorflow==2.4.*  # Gets the CPU version of TF (change to tensorflow-gpu for GPU)
  - pytorch-cpu==1.10.* # Gets the CPU version of TF (change to pytorch for GPU)
  
  # Related to molecular property prediction
  - openbabel==3.*
  - rdkit==2020.03.*
  - ase==3.21.*
  - mordred==1.2.*
  
  # Related to Bayesian
  - corner==2.1.*
  - kombine==0.8.*
  - pymatgen==2020.12.31
  
  # Related to optimal experimental design
  - modal==0.4.*
  - networkx
  - torchani==2.2.2
  - modal==0.4.*
  
  # Utilities
  - tqdm
  - jupyterlab
  - psutil
  - pytest
  - pip
  - pip:
    - git+https://github.com/qmlcode/qml.git@develop
    - dscribe==0.4.0
    - matminer==0.6.*