Reproducible Code for Spherical Principal Curve
Python 3.*
numpy 1.13.3 +
torch 1.1.0 +
pandas
openpyxl
Check Final Code.IPYNB (install jupyter notebook and execute 'jupyter notebook' in CMD)
- utils.py: function modules
- principal_curves.py: codes for principal curve fitting
- main.py: For Earthquake data (fitted curve is saved)
python main.py -q 0.05 -n 500 -e True -i False
Note, q : neighborhood ratio, n : points of a curve
e : ours (True) or Hauberg's(False), i : intrinsic (True) or extrinsic (False)
- silmulation.py: simulation data experiment (excel files are saved):
python simulation.py -q 0.05 -n 100