pyALPS: How to calculate the entanglement_spectrum?

[Catch1123]

I have added
ground_parms['MEASURE[Renyi2]' ] = 1
ground_parms['entanglement_spectra' ] = str(self.L/2)

into my ground_parms to calculate the spectrum.

1. is it right? what dose it return? the singular values? or the entanglement spectra itself? say \mu_i = -2*ln(\lambda_i)?
2. And why my result, the spin-1 finite(L =32) Heisenberg chain (with bond = 800, sweep = 15 and the energy of ground state is -44.84862034/32 = -1.401519385625 close to the exact result −1.401484039...) can not find the even-degeneracy of the data calculated by the ground_parms['entanglement_spectra' ] = str(self.L/2-1) method?

i'm confiused! please help! Thanks very much!

[egull]

@afeiguin my guess is that you're the best person to give some guidance. This is likely from the DMRG code?

I would recommend starting with a tutorial (did you get those to work) and gradually changing parameters so you are always sure you know what you're getting.

[Catch1123]

Thanks so much for your reply！ yes, It's from the mps_optim algorithm. I'm confused because the parameter
ground_parms['entanglement_spectra' ] = str(self.L/2-1) can not be found in the tutorial, therefore I don't have any reference about this parameter to work with.

And my code is also very simple:

import pyalps
from collections import OrderedDict
from copy import deepcopy
import os
def auto_dir(file_path):
dir_path = os.path.dirname(file_path)
if not os.path.exists(dir_path):
os.makedirs(dir_path)
class Spectra:
def init(self, L, Sz_total, num_of_states):
self.L = L
self.Sz_total = Sz_total
self.num_of_states = num_of_states
self.path = os.path.join('Logs', 'L_{}'.format(self.L), 'Spectra',
'Sz{}_Numstates{}'.format(self.Sz_total, self.num_of_states))
def model(self):
model = OrderedDict()
model['LATTICE'] = 'open chain lattice'
model['MODEL'] = 'spin'
model['CONSERVED_QUANTUMNUMBERS'] = 'N,Sz'
model['local_S'] = '1.0'
model['J'] = 1.0
model['MAXSTATES'] = 100 + 25 * self.L
# if self.Sz_total != -1:
# model['Sz_total'] = self.Sz_total
model['Sz_total'] = self.Sz_total
return model
def params(self):
auto_dir(os.path.join(self.path, "log.txt"))
model = self.model()
ground_parms = deepcopy(model)
ground_parms['L'] = self.L
ground_parms['SWEEPS'] = 15
ground_parms['NUMBER_EIGENVALUES'] = self.num_of_states
ground_parms['MEASURE[Renyi2]' ] = 1
ground_parms['entanglement_spectra' ] = str(self.L/2)
ground_parms['MEASURE_LOCAL[local magnetization]'] = 'Sz'
input_file = pyalps.writeInputFiles(os.path.join(self.path, "ground"), [ground_parms])
return input_file
def run(self):
input_file = self.params()
res = pyalps.runApplication('mps_optim', input_file)
def load(self):
return pyalps.loadEigenstateMeasurements(pyalps.getResultFiles(prefix=os.path.join(self.path, "ground")))

[egull]

OK. This is the mps code. It looks like there is nobody active anymore who knows this particular code. We will likely deprecate it since we can no longer provide support. We will discuss steps this Friday at our weekly meeting.