numpy == 1.19.5
torch == 1.8.1
scikit-learn==0.24.2
matplotlib==3.4.2
pandas == 1.2.4
Edit genNpArrays.py
- Edit the variable interested_vars to change what characteristics to use in the dataset
- Use the transform and naive keyword arguments to determine if a transformation will be applied to mu_hit_sim_phi
-
- transform=True, naive=False: Principle direction transformation. (Needs to be checked, do not use for the time being)
-
- transform=True, naive=True: mu_hit_sim_phi replaced with the two variables mu_hit_cos_phi and mu_hit_sin_phi
- Use the comment keyword argument to add additional specifications to the default naming
Run the following commands:
cd Working python main.py
Comments:
- Be sure to load the correct datasets.
- A torch tensorboard file is generated for every training and stored in that training's directory in retrain_runs