experiment.py

#!/usr/bin/env python3

# SPDX-License-Identifier: MIT
# SPDX-FileCopyrightText: Copyright (c) 2023 QOTLabs

import logging
from optparse import OptionParser
from sys import stdout

import numpy as np
from numpy import array
from scipy.optimize import minimize
from math import pi

from vqe import *
from spsa import minimize_spsa

# Option parser
parser = OptionParser()
parser.add_option('-f', '--fake', action='store_true', default=False,
        help='perform simulations instead of real experiment. Not set on default')
parser.add_option('-l', '--log', default='vqe.dat',
        metavar='FILE', help='write output to FILE. Default "%default"')
parser.add_option('-t', '--exposure-time', type='int', default=3000,
        metavar='TIME', help='set exposure time to TIME milliseconds. Default %default')
parser.add_option('-N', '--iterations', type='int', default=150,
        metavar='ITER', help='number of iterations for optimization algorithm. Default %default')
parser.add_option('-i', '--intensity', type='float', default=15000,
        metavar='HERTZ', help='intensity in Hertz that is used for simulations. Default %default')
parser.add_option('-v', '--verbosity', default='warn',
        choices=['debug', 'info', 'warn', 'error'],
        metavar='LEVEL', help='set verbosity level of additional messages to LEVEL. Available: debug, info, warn, error. Default %default')
parser.add_option('-m', type='float', default=0,
        metavar='VAL', help='field parameter in Schwinger Hamiltomian. Default %default')
parser.add_option('-n', '--noise', type='float', default=0,
        metavar='STRENGTH', help='specify decoherence noise strength for simulations. Default %default')
parser.add_option('-c', '--noise-channels', type='int', default=1,
        metavar='CHANNELS', help='specify the number of channels for applying noise. Default %default')
(options, args) = parser.parse_args()

options.verbosity = verbosity = {
        'debug':logging.DEBUG,
        'info':logging.INFO,
        'warn':logging.WARNING,
        'error':logging.ERROR
        }[options.verbosity]

# Hamiltonians
ZZ = {'hh':(-1,1,1,-1)} # -ZZ
ZZXX = {'hh':(-1,1,1,-1), 'dd':(-2,0,0,2)} # -ZZ-XI-IX
ZZYY = {'hh':(-1,1,1,-1), 'rr':(-2,0,0,2)} # -ZZ-YI-IY
XZZX = {'dh':(1,-1,-1,1), 'hd':(1,-1,1,-1)} 
def schwinger(m):
    return {'hh':(1,-m,1+m,2), 'dd':(1,-1,-1,1), 'rr':(1,-1,-1,1)} # 1+XX+YY+0.5(-ZI+ZZ+mIZ-mZI)

# Initialization
logging.basicConfig(format='%(asctime)-15s [%(levelname)s] %(message)s', level=options.verbosity)
setup = Setup(fake=options.fake)
setup.set_exposure_time(options.exposure_time)
setup.intensity = options.intensity # Intensity in Hertz for simulations
setup.noise = options.noise
setup.noise_channels = options.noise_channels
setup.efficiencies = [1.66806, 1.4648, 0.789505, 0.687237]
setup.retardances = [pi/2, 2*pi*0.4995, 2*pi*0.2493, 2*pi*0.5129, 2*pi*0.2198, 2*pi*0.4799] # Waveplace retardances
# Hamiltonian evaluator
H = MeanValue(setup, schwinger(options.m))

logfile = open(options.log, 'w')
log_header(setup, H, stdout)
log_header(setup, H, logfile)

# Optimization
def callback_func(x):
    result = H(x)
    log_data(stdout, x, result)
    log_data(logfile, x, result)
def target_func(x):
    return H(x)[0]
x0 = [1,0,0,0,0,0]
#x0=np.random.uniform(0,2*pi,6)
m = options.m
result = minimize_spsa(target_func, callback=callback_func, x0=x0, maxiter=options.iterations,
        a0=0.05/(0.2*abs(m)+1), af=0.005/(0.2*abs(m)+1), b0=0.1, bf=0.002)

# Order operator
O = MeanValue(setup, {'hh':(0,0,1,0)})
log_data(stdout, result.x, O(result.x))
log_data(logfile, result.x, O(result.x))