Two Step Locality Preserving Projections (TS-LPP)
Python >= 3.7
numpy >= 1.16.2
scikit-learn >= 0.20.3
scipy >= 1.4.1
matplotlib >= 2.2.3
Download or clone the github repository, e.g. git clone https://github.com/rtmr/TS-LPP
n_components: int (The reduced dimension in the final step)
knn: int (The number for neighbor graph in LPP)
n_clusters: int (Number of target clusters)
Target dataset is set to the feature matrix without label information. (Target.csv contains the LAAF descriptors for solid, liquid, and amorphous states in silicon.)
python TS-LPP.py
This project is licensed under the terms of the MIT license.