Ranking of Pareto solutions based on projection free-energy
Python >= 3.7
rho: int (parameter in augmented weighted Tchebycheff)
num_top: int (number of proposals)
wind: float (window of weights)
Tstar: float (value of Tstar)
The target dataset is prepared as data.csv.
The example is (two-dimensional property space) as follows.
| name | objective1 | objective 2 |
|---|---|---|
| Sc2O3 | 3.75 | 0.672 |
| GeO2 | 3.05 | 0.518 |
| : | : | : |
In present, two- and three-dimensional property space can be targetted.
python rppf.py
This project is licensed under the terms of the MIT license.