# 'legend.fontsize': 14, # 8
# 'font.size': 14, # 8
# 'xtick.labelsize': 14, # 8
# 'ytick.labelsize': 14, # 8
# 'lines.markersize': 16, # 8
# 'savefig.bbox': 'tight', # tight
# 'savefig.pad_inches': 0.1, # 0.1
# 'text.usetex': True,
# 'font.family': 'serif',
# 'text.latex.preamble' : [r'\usepackage{amsmath}'],
# 'figure.figsize': (10, 7)} # (3.5, 2.5) default
# rcParams.update(plot_params)
i=10
stri = str(i)
IO.go_to_data_parent('interpolatedboson/a'+stri)
IO.add_to_path()
state = IO.get_state()
Ls = range(9, 11, 1)
L_min, L_max = min(Ls), max(Ls)
shift_func = lambda L: 0.1 if L==9 else -0.15
#shift_func = lambda L: 0.2*(L-L_max)/(L_max - L_min)
#shift_func = lambda L: 0
params = IO.Params.kw('LKN', L=Ls, K=[0], N=[0])
for p in params:
run.create_edge_spectrum(state, p, overwrite=False)
spectrum = pp.load_edge_spectrum(state, Ls)
spectrum = pp.shift_boson_number(spectrum)
# coding: utf-8
from PEPS import mps_tools
from __future__ import division
import math
import numpy as np
from PEPS import IO_helper as IO
IO.go_to_data_parent('softcoreboson')
IO.add_to_path()
state = IO.get_state()
mps = mps_tools
mps.example_SPT_calc()
L=3
p=2
edge_d=2
#lcst = get_left_canonical_site_tensor(statename='hardcoreboson', L=L)
gcst, Lambda, V_basis_change = mps.get_canonical_site_tensor(statename='hardcoreboson', L=L)
p_dim, bond_dim, _ = gcst.shape
reload(mps)
L=3
p=2
edge_d=2
#lcst = get_left_canonical_site_tensor(statename='hardcoreboson', L=L)
gcst, Lambda, V_basis_change = mps.get_canonical_site_tensor(statename='hardcoreboson', L=L)
p_dim, bond_dim, _ = gcst.shape
#print rhoL
XHC = np.array([[0, 1], [1, 0]])
ZHC = np.array([[1, 0],[0,-1]])
ZSC = np.array([[1, 0, 0, 0],[0,-1, 0, 0],[0,0, 1, 0],[0,0, 0, -1]])
Id = np.identity(p_dim)
cax10 = divider10.append_axes("right", size="5%", pad=0)
divider11 = make_axes_locatable(ax11)
cax11 = divider11.append_axes("right", size="5%", pad=0)
L = 8
radius1 = radius2 = 5.0
hexradius1 = hexradius2 = 4.2
# radius1 = radius2 = 14.0
# hexradius1 = hexradius2 = 13.5
circle_size = "radius"
## Left two plots for softcoreboson
IO.go_to_data_parent("softcoreboson", parent="Data//FBI-TM")
params = IO.Params.kw("LKN", L=[L], K=[0], N=[0])
param = next(iter(params))
corrmap, corrmap_string = plot.load_combined_corrmap(param, ("bdag", "b"))
hexcorrmap = hexplot.convert_zigzag_dict_to_hex(
corrmap, L, [p for p, dist in hexplot.Hex(-1, 2).disc(radius=radius1) if p.sub != "C"]
)
hexcorrmap = {k: np.real(v).item() for k, v in hexcorrmap.items()}
p = hexplot.hex_circle_plot(
hexcorrmap,
hexlattice_radius=hexradius1,
cmap=hexplot.cmap_seq,
ax=ax00,
circle_size=circle_size,
scale=1,
import numpy as np
from matplotlib import pyplot as plt
from DataAnalysis import PEPS_plots as pp
from PEPS import spectra_analyzer as spec
from PEPS.CylinderPEPS import run
from PEPS import analysis
from PEPS import IO_helper as IO
import sfig
i=10
stri = str(i)
IO.go_to_data_parent('interpolatedboson/a'+stri)
IO.add_to_path()
state = IO.get_state()
Ls = range(2, 11)
params = IO.Params.kw('LKN', L=Ls, K=[0], N=[0])
for p in params:
run.create_edge_spectrum(state, p, overwrite=False)
def entanglement_entropy(es):
return sum(-v * np.log2(v) for v in es if not np.isclose(v, 0))
entropy = []
for p in params:
spec = run.load_edge_spectrum(state, p, evecs=False)
es = [spec_pnt[1] for spec_pnt in spec]
entropy.append(entanglement_entropy(es))
entropy = np.real(entropy)
from matplotlib.backends.backend_pdf import PdfPages
from DataAnalysis import data
from DataAnalysis import PEPS_plots as pp
import numpy as np
from matplotlib import pyplot as plt
from PEPS import IO_helper as IO
from PEPS import spectra_analyzer as spec
from PEPS.CylinderPEPS import run
show_fit_details = True
show_legend = True
############ Go to state data
i = 10
stri = str(i)
IO.go_to_data_parent("interpolatedboson/a" + stri)
IO.add_to_path()
state = IO.get_state()
Ls = range(2, 11, 2)
L_min, L_max = min(Ls), max(Ls)
shift_func = lambda L: 0.3 * (L - L_max - 1) / (L_max - L_min)
# if needed create edge spectrum
params = IO.Params.kw("LKN", L=Ls, K=[0], N=[0])
for p in params:
run.create_edge_spectrum(state, p, overwrite=False)
# load and prepare spectrum
spectrum = pp.load_edge_spectrum(state, Ls)
spectrum = pp.shift_boson_number(spectrum)
spectrum.add_as_prop(pp.energy, val="val")
from DataAnalysis import data
from DataAnalysis import PEPS_plots as pp
import numpy as np
from matplotlib import pyplot as plt
from PEPS import IO_helper as IO
from PEPS import spectra_analyzer as spec
from PEPS.CylinderPEPS import run
from collections import OrderedDict
show_fit_details = False
show_legend = True
############ Go to state data
i=10
stri = str(i)
IO.go_to_data_parent('interpolatedboson/a'+stri)
IO.add_to_path()
state = IO.get_state()
Ls = range(2, 11, 2)
L_min, L_max = min(Ls), max(Ls)
shift_func = lambda L: 0.3*(L-L_max-1)/(L_max - L_min)
# if needed create edge spectrum
params = IO.Params.kw('LKN', L=Ls, K=[0], N=[0])
for p in params:
run.create_edge_spectrum(state, p, overwrite=False)
# load and prepare spectrum
spectrum = pp.load_edge_spectrum(state, Ls)
spectrum = pp.shift_boson_number(spectrum)
spectrum.add_as_prop(pp.energy, val='val')
# coding: utf-8
from __future__ import division
from DataAnalysis import data
from DataAnalysis import PEPS_plots as pp
import numpy as np
from matplotlib import pyplot as plt
from PEPS import IO_helper as IO
from PEPS import spectra_analyzer as spec
from PEPS.CylinderPEPS import run
#IO.go_to_data_parent('rotated_softcoreboson')
state = IO.get_state()
params = IO.Params.kw('LKN', L=[1,2, 3], K=[0], N=[0])
for p in params:
run.create_edge_spectrum(state, p)
reload(run)
params = IO.Params.kw('LKN', L=[1,2, 3], K=[0], N=[0])
for p in params:
run.create_edge_spectrum(state, p)
def rotated_state_params(Ls):
params = data.pdict.kw('LKN', L = Ls, K = range(max(Ls)), N = range(-max(Ls), max(Ls)+1))
K_lt_L = lambda p: p['K']<p['L']
N_le_L = lambda p: -p['L'] <= p['N'] <= p['L']
params.add_constraints([K_lt_L, N_le_L])
return params
Ls = [1, 2, 3]
L_min = min(Ls)
L_max = max(Ls)
params = rotated_state_params(Ls)
spectrum = data.ParametrizedDataSet(params)
from matplotlib import rcParams
rcParams.update({'figure.figsize':(5.3, 4)})
Ls = range(2, 9)
scale = 0
L_min = min(Ls)
L_max = max(Ls)
# from PEPS.CylinderPEPS import run
# run.go_diag_sectors(8, 8, sectors=[[0,1], [1,1]])
# run.go_diag_sectors(5, 5, sectors=[[0,1], [1,1]])
# run.go_diag_sectors(5, 5, sectors=[[0,1], [1,1]], overwrite=True)
#pp.IO.go_to_data_parent('interpolatedboson/a0')
IO.go_to_data_parent('softcoreboson')
IO.add_to_path()
state = IO.get_state()
spectrum = pp.load_transfer_matrix_spectrum(state, Ls)
pp.shift_and_scale(spectrum, scale)
spectrum = spectrum.filter([lambda p:(0<=p['K']<4 and 0<=p['N']<4)], [lambda band:band<=1], band='band')
spec_points = spectrum.collect(['K', 'N', 'band'], 'L', 'E') # ParametrizedDataSet with K, N, band as parameters
spec_points.plot_toolbox = pp.data.SpectraPlotTools()
spec_points.plot_toolbox.markers = ['1', '2', '3', '4']
spec_points.update_plot_args(title='Transfer Matrix Spectrum',
xlabel='K', ylabel='Inverse Correlation Length -log(|T|) = 1/$\\xi$',
xlim=[-0.5, 2.1], ylim=[-0.1, 2.1], ls = ':',
markersize=8)
from __future__ import division
from matplotlib import pyplot as plt
import numpy as np
from DataAnalysis import PEPS_plots as pp
from PEPS import IO_helper as IO
from PEPS.CylinderPEPS import run
import sfig
ev_to_corr = lambda e: -1/np.log(np.abs(e))
phase = lambda e: 'e^(i pi {})'.format(IO.round_sig(np.angle(e)/np.pi, 2))
def corrbound_data(state, Ls):
ans = {}
for L in Ls:
tmspec = run.load_transfer_matrix_spectrum(state, L)
if tmspec:
norm = tmspec[(0, 0)][0]
ev1 = tmspec[(0, 1)][0]/norm
if L>1:
symev00 = tmspec[(0, 0)][1]/norm
symev10 = tmspec[(1, 0)][0]/norm #K=+-1, N=0 ev is bigger than K=0, N=0 sometimes
print abs(symev00), abs(symev10)
symev = symev00 if abs(symev00)>abs(symev10) else symev10
#symev = symev00
else:
symev00 = tmspec[(0, 0)][1]/norm
symev = symev00
ans[L] = ev_to_corr(ev1), phase(ev1), ev_to_corr(symev), phase(symev)
return ans
IO.go_to_data_parent('softcoreboson', parent='Data//FBI-TM')
# coding: utf-8
get_ipython().system(u'explorer .')
from __future__ import division
from DataAnalysis import data
from DataAnalysis import PEPS_plots as pp
import numpy as np
from matplotlib import pyplot as plt
from PEPS import IO_helper as IO
from PEPS import spectra_analyzer as spec
from PEPS.CylinderPEPS import run
IO.go_to_data_parent('armchair_softcoreboson')
state = IO.get_state()
params = IO.Params.kw('LKN', L=[1,2, 3], K=[0], N=[0])
from PEPS import spectra_analyzer as spec
def armchair_state_params(Ls):
params = data.pdict.kw('LKN', L = Ls, K = range(max(Ls)), N = range(-max(Ls), 2*max(Ls)+1))
K_lt_L = lambda p: p['K']<p['L']
N_le_2L = lambda p: -p['L'] <= p['N'] <= 2*p['L']
params.add_constraints([K_lt_L, N_le_2L])
return params
Ls = [1, 2, 3]
params = armchair_state_params(Ls)
spectrum = data.ParametrizedDataSet(params)
for L in Ls:
edge_spectrum_list = run.load_edge_spectrum(state, data.OrderedDict(zip(['L', 'K', 'N'], [L, 0, 0])))
edge_spectrum_dict = spec.collect_spectrum_points(edge_spectrum_list)
for K, N in edge_spectrum_dict:
for indx, val in enumerate(edge_spectrum_dict[(K, N)]):
spectrum.append(data.DataPoint(L=L, K=K, N=N, val=val, band=indx))
Ls = [1, 2, 3]
params = armchair_state_params(Ls)
def go(L):
#with PdfPages('L_10_all_mom_10.pdf') as pdf:
###########################################################
#get state
i=10
stri = str(i)
IO.go_to_data_parent('interpolatedboson/a'+stri)
IO.add_to_path()
state = IO.get_state()
Ls = [L]
L_min, L_max = min(Ls), max(Ls)
shift_func = lambda L: 0
###########################################################
#make edge spectrum if necessary
params = IO.Params.kw('LKN', L=Ls, K=[0], N=[0])
for p in params:
run.create_edge_spectrum(state, p, overwrite=False)
###########################################################
#load and prepeare edge spectrum
spectrum = pp.load_edge_spectrum(state, Ls)
spectrum = pp.shift_boson_number(spectrum)
spectrum.add_as_prop(pp.energy, val='val')
data_name = 'energy'
pp.shift_and_scale(spectrum, scale=0) #just shift
data_name = 'E'
all_points = list(spectrum.points())
spec_points = spectrum.collect(['K', 'N', 'band'], 'L', data_name)
spec_points = spec_points.filter([lambda p: p['K']<=L/2])
spec_points.plot_toolbox = data.SpectraPlotTools()
spec_points.plot_toolbox.markers = ['1', '2', '3', '4']
spec_points.update_plot_args(title='', xlabel='K', ylabel='Energy',
xlim=[-0.5, 12], ylim=[-0.4, 6])
#even_colors = ['k', 'b', 'g', 'r', 'gold', 'orange', 'orangered', 'y','m', 'lime', 'c']
color_pairs = [('k', 'k'), ('b', 'c'),('g', 'lime'),('r', 'm'),('orange', 'orangered'),('gold', 'y')]
colors = [c for cp in color_pairs for c in cp]
def color_func(N):
x=0
if N<0:
x=1
if N%2==0:
return abs(N)+x
else:
return abs(N)+1+4+x
spec_points.plot_toolbox.colors = colors
spec_points.plot_toolbox.color_func = color_func
###########################################################
#make plot
if makelegend:
xlim=[-0.5, np.floor(L/2)+1.1]
else:
xlim=[-0.5, np.floor(L/2)+.5]
fig, spec_ax = spec_points.plot_toolbox.spec_plot(spec_points, data_name,
shift_func=shift_func, copy_points=True,xlim=xlim, legend=None)
# if i!=10 and i!=0:
# plt.title('Entanglement Spectrum for b=0.'+stri)
# elif i==10:
# plt.title('Entanglement Spectrum for soft-core state')
# elif i==0:
# plt.title('Entanglement Spectrum for hard-core FBI')
if L==10:
plt.ylabel('Entanglement Energy')
if L==9:
plt.ylabel('')
ax = plt.gca()
Ks = np.arange(0, np.floor(L/2+1))
ax.set_xticks(Ks)
if L==10:
ax.set_xticklabels(['0', '$\pi/5$', '$2\pi/5$', '$3\pi/5$', '$4\pi/5$', '$\pi$'])
if L==9:
ax.set_xticklabels(['0', '$2\pi/9$', '$4\pi/9$', '$6\pi/9$', '$8\pi/9$'])
#ax.set_xticklabels(['$2\pi'+str(K)+'/'+str(L)+'$' if K in [1for K in Ks])
ax.yaxis.tick_left()
ax.xaxis.tick_bottom()
if makelegend:
handles, labels = ax.get_legend_handles_labels()
hl = sorted(zip(handles, labels),
key= lambda x: abs(int(x[1])))
handles2, labels2 = zip(*hl)
def label_helper(string_representing_integer):
i = int(string_representing_integer)
if not i%2:
if i<=0:
return str(i//2)
else:
return '+'+str(i//2)
else:
if i<=0:
return str(i)+'/2'
else:
return '+'+str(i)+'/2'
labels3 = [label_helper(l) for l in labels2]
if L==9:
legendloc = (0.8, 0.15)
if L==10:
legendloc = (0.825, 0.15)
leg=ax.legend(handles2, labels3, loc = legendloc, numpoints=1, frameon=False, labelspacing=0.25)
