Huayi Providers

Basic Use

Download packages to your working directory

git clone https://github.com/coding-xyz/huayi-providers huayi_providers

Import backend

from huayi_providers.fake_{backend_name} import *
Fake{backend_name}()    # for V1 backend
Fake{backend_name}V2()  # for V2 backend

Generate noise model

from qiskit_aer.noise.noise_model import NoiseModel
noise_model = NoiseModel.from_backend(Fake{backend_name}())

Build backends from .csv files

from huayi_providers.backend_build import build_from_file
c = build_from_file(backend_name,
                    backend_version,
                    qubits_data,
                    gates_data)

File Structure

Experiment data

Qubits properties are Stored in ./data/qubits_data_{backend_name}.csv and gates information in ./data/gates_data_{backend_name}.csv.

qubits_data_{}.csv contains the information of qubits, including

• T1 time (ms)
• T2 time (ms)
• frequency (MHz)
• readout error rate
• Probability of finding 0 when prepared in 1
• Probability of finding 1 when prepared in 0
• readout length (us)

All information should be accompanied with the measurement date and time.

gates_data_{}.csv contains the information of gates, including

• qubits
• gate type
• error rate
• length
• gate name (optional)

The gate error and length are measured from experiment, and should be accompanied with the measurement date and time.

Backend files

from backend_build import build_from_file to generate the files for the backend.

• ./fake_{backend_name}/
  • __init__.py
  • fake_{backend_name}.py
  • props_{backend_name}.json
  • conf_{backend_name}.json
  • defs_{backend_name}.json (if Pulse Backend is applied, TODO)

Dictionaries of the backend properties and configurations

props_huayi.json

{'backend_name': 'fakehuayi',
 'backend_version='0.0.1',
 'last_update_date': now_time(),
 'qubits': [q1, q2, ...],
 'gates': [g1, g2, ...],
 'general': []}

qubit info (q):

 [{'date': ['T1_date'], 'name': 'T1', 'unit': 'ms', 'value': ['T1']},
  {'date': ['T2_date'], 'name': 'T2', 'unit': 'ms', 'value': ['T2']},
  {'date': ['frequency_date'], 'name': 'frequency', 'unit': 'MHz', 'value': ['frequency']},
  {'date': ['readout_error_date'], 'name': 'readout_error', 'unit': '', 'value': ['readout_error']},
  {'date': ['prob_meas0_prep1_date'], 'name': 'prob_meas0_prep1', 'unit': '', 'value': ['prob_meas0_prep1']},
  {'date': ['prob_meas1_prep0_date'], 'name': 'prob_meas1_prep0', 'unit': '', 'value': ['prob_meas1_prep0']},
  {'date': ['readout_length_date'], 'name': 'readout_length', 'unit': 'us', 'value': ['readout_length']}]

gate info (g):

{'qubits': ['qubits'],
 'gate': ['gate'],
 'parameters': [{'date': ['error_date'],
                 'name': 'gate_error',
                 'unit': '',
                 'value': ['gate_error']},
                {'date': ['length_date'],
                 'name': 'gate_length',
                 'unit': 'ms',
                 'value': ['gate_length']}],
 'name': ['name']}

conf_huayi.json

Dictionary of the Noise Model

The keys in NoiseModel are

• basis gates ({'id', 'x', 'sx', 'rz', 'cx'})
• noise instructions ({'id', 'x', 'sx', 'cx', 'measure', 'reset', ''})
• noise qubits ({0, ... , 26})
• default quantum errors (set None by default)
• default readout errors (set None by default)
• local quantum errors
• local readout errors
• custom noise passes (set None by default)

local readout errors

noise_model._local_readout_errors = {(0,):ReadoutError, 
                                     (1,):ReadoutError, 
                                     (2,):ReadoutError, ...}

ReadoutError is essentially a matrix, the off-diagonal terms corresponds to prob_meas0_prep1 and prob_meas1_prep0

If prob_meas0_prep1 and prob_meas1_prep0 are given, readout_error will be ignored.

local quantum errors

The errors of gate applied to indices ind = (i,) for one-bit gate, (i,j) for two-bit gate are stored in

noise_model._local_quantum_errors[gate][ind] = {
  '_id': xxxx,
  '_probs': list_of_probs,
  '_circs': list_of_circs,
  '_qargs': None,
  '_op_shape': OpShape(num_qargs_l, num_qargs_r)}

_circs includes all possible matrices with the same shape of the corresponding gate

_probs is the corresponding probability of each circ

For one-bit gate, _circ includes I, X, Y, Z. The probs of nonidentity matrices are gate_error/2. gate_error in .csv file refers to the measurement of 1-(I+Z).

For two-bit gate, _circ includes II, IX, IY, IZ, XI, XX, XY, XZ, YI, YX, YY, YZ, ZI, ZX, ZY, ZZ. The probs of nonidentity matrices are gate_error/12. The gate_error in .csv file refers to the measurement of 1-(II+IZ+ZI+ZZ).