Prodce_Data_Sparse.py

from quspin.operators import hamiltonian  # Hamiltonians and operators
from quspin.basis import spin_basis_1d  # Hilbert space spin basis
import numpy as np  # generic math functions
import scipy.sparse
import h5py
from keras.utils.io_utils import HDF5Matrix
import matplotlib
matplotlib.use('Agg')
import pylab as plt


def Hamilt_qspin(Jxy, Jzz_0, hmax, N):
    #basis = spin_basis_1d(L,pauli=False)
    no_checks = {"check_herm": False, "check_pcon": False, "check_symm": False}
    basis = spin_basis_1d(
        N, pauli=False, Nup=N // 2)  # zero magnetisation sector
    #pauli false indicates that we use 1/2 sigma convention
    #Nup is the number of spins pointing up.
    hz = np.random.uniform(-hmax, hmax, N)
    # define operators with OBC using site-coupling lists
    J_zz = [[Jzz_0, i, (i + 1) % N] for i in range(N)]  # PBC
    J_xy = [[Jxy / 2.0, i, (i + 1) % N] for i in range(N)]  # PBC
    h_z = [[hz[i], i] for i in range(N)]

    # static and dynamic lists
    static = [["+-", J_xy], ["-+", J_xy], ["zz", J_zz]]
    disorder_field = [["z", h_z]]
    dynamic = []
    # compute the time-dependent Heisenberg Hamiltonian
    H_XXZ = hamiltonian(
        static, dynamic, basis=basis, dtype=np.float64, **no_checks)
    Hz = hamiltonian(
        disorder_field, [], basis=basis, dtype=np.float64, **no_checks)
    Htot = H_XXZ + Hz
    return Htot


def statistics(n, E):
    delta_n = (E[n] - E[n - 1])
    delta_np1 = (E[n + 1] - E[n])
    #print(delta_n, delta_np1, min(delta_n, delta_np1)/max(delta_n, delta_np1))
    return min(delta_n, delta_np1) / max(delta_n, delta_np1)


def find_epsilon_index(Energies, epsilon):
    # find epsilon = 0.5
    # 0.5 = (E-Emax)/(Emin-Emax) --> E = 0.5*(Emin+Emax)
    targetE = epsilon * (Energies[0] - Energies[-1]) + Energies[-1]
    return np.abs(np.array(Energies) - targetE).argmin()


def mean_energy(samples, energies, hmax, epsilon):
    epsilon_index = find_epsilon_index(energies, epsilon)
    deltas = []
    for n in range(
            max(1, epsilon_index - 25),
            min(len(energies) - 1,
                epsilon_index + 25)):  #for n in range(1,2**N-1): #
        deltas.append(statistics(n, energies))
    mean = np.mean(deltas)
    return mean, epsilon_index


def calc_samples(epsilon, hmax, margin, stat):
    H = Hamilt_qspin(Jxy, Jzz_0, hmax, L)
    # Harray = H.toarray()
    # H_sparse = scipy.sparse.csr_matrix(Harray)
    # print(H_sparse.shape)
    emin, emax = H.eigsh(
        k=2, which="BE", maxiter=1E4, return_eigenvectors=False)
    e = epsilon * (emin - emax) + emax
    # energies, states = scipy.sparse.linalg.eigsh(H_sparse, k=2*margin+2, sigma=e, maxiter=1E4, return_eigenvectors= True)
    energies, states = H.eigsh(
        k=2 * margin + 2, sigma=e, maxiter=1E4, return_eigenvectors=True)
    states = states.T
    sort_indices = energies.argsort()
    energies = energies[sort_indices]
    states = states[sort_indices]
    kk = 2 * margin
    r = 0
    for i in np.arange(1, kk + 1):  #why cutting first and last off?
        delta_n = energies[i] - energies[i - 1]
        delta_n1 = energies[i + 1] - energies[i]
        r = r + min(delta_n, delta_n1) / max(delta_n, delta_n1)
    r = r / (2 * margin)
    for i in np.arange(1, kk + 1):
        #psi = np.asarray(i)
        #psi = psi.reshape(psi.shape[1],)
        state_list.append(states[i])
        if hmax < 2.0:
            labels.append([1, 0])
        else:
            labels.append([0, 1])
    return r


def save_H5(filename):
    f = h5py.File(filename, 'w')
    # Creating dataset to store features
    X_dset = f.create_dataset(
        'my_data', (len(labels), state_list[0].shape[0]), dtype='f')
    X_dset[:] = state_list
    # Creating dataset to store labels
    y_dset = f.create_dataset('my_labels', (len(labels), 2), dtype='i')
    y_dset[:] = labels
    f.close()


L = 10  # system size
Jxy = 1.0  # xy interaction
Jzz_0 = 1.0  # zz interaction

samples = 10
margin = 25
labels = []
state_list = []
labels_fullspec = []
state_list_fullspec = []
labels_2D = []
state_list_2D = []
folder = 'Data/'

source = True
target = True
evaluation = True

epsilon = 0.5
hmax_list = np.linspace(0.5, 5.0, 20)
hmax_list = [5.0]
liste = []
statfolder = 'SPARSE_STATISTICS/'

#for hmax in hmax_list:
#    stat = 0.0
#    for i in range(samples):
#        print(hmax, i)
#        stat += calc_samples(epsilon, hmax, margin, stat)
#    liste.append(stat/samples)

if source:
    # SOURCE
    labels = []
    state_list = []
    hmax_list = np.append(np.linspace(0.8, 0.85, 2), np.linspace(5.1, 5.2, 2))
    epsilon_list = epsilon_list = np.linspace(0.1, 0.9, 9)
    filename = folder + '00_SPARSE_SOURCE_N' + str(L) + '.h5'
    for epsilon in epsilon_list:
        liste = []
        for hmax in hmax_list:
            stat = 0.0
            for i in range(samples):
                print('SOURCE, ', 'epsilon:', epsilon, 'hmax:', hmax, i)
                stat = calc_samples(epsilon, hmax, margin, stat)
            liste.append(stat / samples)
        np.save(
            statfolder + 'SPARSE_SOURCE_statistics_N_' + str(L) + 'epsilon_' +
            str(epsilon), liste)
    save_H5(filename)

if target:
    # TARGET
    hmax_list = np.linspace(0.9, 5.0, 20)
    epsilon_list = [0.5]
    filename = folder + '00_SPARSE_TARGET_N' + str(L) + '.h5'
    for epsilon in epsilon_list:
        liste = []
        for hmax in hmax_list:
            stat = 0.0
            for i in range(samples):
                print('TARGET, ', 'epsilon:', epsilon, 'hmax:', hmax, i)
                stat = calc_samples(epsilon, hmax, margin, stat)
            liste.append(stat / samples)
        np.save(
            statfolder + 'SPARSE_TARGET_statistics_N_' + str(L) + 'epsilon_' +
            str(epsilon), liste)
    save_H5(filename)

for samples in [10, 100, 1000]:
    if evaluation:
        # EVALUATION SET
        labels = []
        state_list = []
        folder = 'SPARSE_EVALUATION_FILES/'
        hmax_list = np.linspace(0.9, 5.0, 20)
        epsilon_list = epsilon_list = np.linspace(0.1, 0.9, 9)
        for epsilon in epsilon_list:
            liste = []
            for hmax in hmax_list:
                stat = 0.0
                labels = []
                state_list = []
                labels_fullspec = []
                for i in range(samples):
                    print('SPARSE_EVALUATE, ', 'epsilon:', epsilon, 'hmax:',
                          hmax, i)
                    stat = calc_samples(epsilon, hmax, margin, stat)
                liste.append(stat / samples)
                filename = folder + str(samples) + '_samp_EVALUATE_N' + str(
                    L) + 'eps' + str(epsilon) + 'hmax' + str(hmax) + '.h5'
                print(len(state_list))
                save_H5(filename)
            np.save(
                statfolder + 'SPARSE_EVALUATION_statistics_N_' + str(L) +
                '_sampels_' + str(samples) + 'epsilon_' + str(epsilon), liste)