Codes for the Numerical Analysis in the paper SymmPI (SymmPI: Predictive Inference for Data with Group Symmetries).
The files named 'supervised/unsup_alpha0.05/0.15_fix/random.py' comprise the codes utilized for simulation studies corresponding to the settings in either supervised or unsupervised learning. These settings are characterized by a significance level of alpha = 0.05 or 0.15 and a fixed or random number of observations per branch. For a comprehensive understanding, refer to Sections 5.6 and 7.6 of the paper. In addition, the file named symmPI.R contains the codes for real data study presented in section 5.7 of the paper.