params.py

import argparse
import torch
import numpy as np
# n = 64                      # signal dimension
# alpha = 10
# m = alpha * n               # number of measurements
# # m = 2
# cplx_flag = 1               # real: cplx_flag = 0;  complex: cplx_flag = 1;
# T = 80                      # number of iterations
# npower_iter = 30            # number of power iterations
# N_train= 100                # Number of training samples
# EPOCHS = 1
# scenario = 4 
# LR = 1e-3

# N = N_train                 # Number of training samples
# mu = 0.8+0.4*cplx_flag      # suggested step for the Wirtinger Flow
# cuda_opt = 1
# if torch.cuda.is_available() & cuda_opt:
#     DEVICE = "cuda"
# else:
#     DEVICE = "cpu"              

# batch = n


# scalar = True
# vector = True
# matrix = True
# tensor = True
# if scenario == 0:
    # scalar = True 
    # vector = False
    # matrix = False
# if scenario == 1:
    # scalar = False 
    # vector = True
    # matrix = False
# if scenario == 2:
    # scalar = False
    # vector = True
    # matrix = True
# if scenario == 3:
    # scalar = False
    # vector = False
    # matrix = True




# print(Tensor)
# parser = argparse.ArgumentParser()
# parser.add_argument("-s", help="Scenario", type=int)
# parser.add_argument("-lr", help="Learning Rate", type=float)
# parser.add_argument("-e", help="epochs", type=int)
# parser.add_argument("-n", help="SNR", type=int)
# parser.add_argument("-c", help="Scenario", type=int)
# parser.add_argument("-tstart", help="T start (Number of unfoldings)", type=int)
# parser.add_argument("-tend", help="T end (Number of unfoldings)", type=int)
# parser.add_argument("-tstep", help="T step (Number of unfoldings)", type=int)
# parser.add_argument("-ntrain", help="N Train", type=int)
# parser.add_argument("-sigsnr", help="Signal SNR", type=int)
# parser.add_argument("-epochs", help="Number of epochs", type=int)
# args = parser.parse_args()
# scenario = args.s
# LR = args.lr   
# EPOCHS = args.e