

# Code
HTE_publication.ipynb  : jupyter notebook
it can be used to re-produce the final model

# Environnment
HTE-env.yml: describe the python environnement that has been used during the modelling


# The dataset files contain these information for each experiment:
- experiment identifier
- the reaction smiles 
- the coupling agent 
- Time
- Temperatute
- % of Conversion %

Note:
- HTE_CV.txt contains 836 experiments used for the cross-validation procedure (using random sampling)
- HTE_Amine_1.txt, HTE_Amine_2.txt, ... contains 836 experiments used for the leave-one-amine out process, each experimnent has been tagged as "Train" or "Test" depending it has to be considered in the training or test set, respectively
- HTE_Final.txt contains 1070 experiments tagged as "Train" or "Test" depending it has to be considered in the training or test set, respectively. All the "train" correspond to the first campaign of experiments, all the "test" correspond to the second campaign of experiments (e.g. "temporal test set")


# Phychem properties used in some of the models
Reactants.properties.txt: contains the physico-chemical properties computed with "oeselma" (propriatory method that cannot be shared) for the different reactants


