Dataset splitting does not work as intended for some cases

choderalab / espaloma

Extensible Surrogate Potential of Ab initio Learned and Optimized by Message-passing Algorithm 🍹https://arxiv.org/abs/2010.01196

https://docs.espaloma.org/en/latest/

MIT License

203 stars 23 forks source link

Dataset splitting does not work as intended for some cases #149

Open kntkb opened 1 year ago

kntkb commented 1 year ago

A data can be left out from the train/val/test dataset when partitioning the data (here). For example a dataset with 64 entries will be partitioned into 51:6:6 entries if they are split into ratios of 80:10:10 which does not sum up to 64 entries.