Reduce number of qubits used in the following cases:

• SPLIT_SLIDE and MERGE_SLIDE: if an arm of the slide/merge only has one path qubit then that qubit could be used directly as the path qubit, instead of adding a new ancilla qubit;
• CAPTURE: don't create a capture ancilla if the attacker is known to be there and there is only one path qubit.

• Add Hamiltonian and charge/spin densities.
• Add frontmatter (run on Colab, download notebook, etc.)
• Used standard install/import scheme.
• Remove unused imports.
• Add a bit more explanatory text.
• Minor typo fixes, changes, etc.

• Migrated device.qubits callsites
  • A None check is added if device is a cirq.Device.
• Migrated
• Removed gateset parameters

The first commit conflicts with https://github.com/quantumlib/ReCirq/pull/306 in that the solution is to check for device.metadata is None rather than requiring GridDevice. Either approach SGTM.

Couldn't test locally because pytket.routing is missing. @mpharrigan does https://github.com/quantumlib/ReCirq/pull/306 fix this?

Pinning the requirements required fixing versions of the following packages: pytket-cirq==0.3 and pytket=0.5.2 due to the latest pytket relying on openfermion~=0.11 and hfvqe relying on openfermion~=1.0.0.

Also the following requirements were added:

• pytest (for efficient test suite running)
• numpy (used in various tests but wasn't explicitly required)
• pip-compile (for automated requirements generation)

This experiment has some code in cirq-core/cirq/experiments, along with a docs notebook. Both have been moved here, into recirq/rabi_oscillations and docs/rabi_oscillations respectively

Mostly file moves and renames. Some Markdown fixes.

Some Sphinx-specific files moved into dev_tools/docs/sphinx/ —just in case you need them and can probably delete from repo when you are ready.

A new Cirq has been released. I will track any outstanding migration issues here.

• [ ] bump cirq version number
• [ ] ERROR: cirq 0.11.1 has requirement cirq-core==0.11.1, but you'll have cirq-core 0.12.0 which is incompatible. ERROR: cirq 0.11.1 has requirement cirq-google==0.11.1, but you'll have cirq-google 0.12.0 which is incompatible. I think this is from openfermion https://github.com/quantumlib/OpenFermion/issues/743. ~~pip install seems to work anyway: the above is a warning.~~ newer pips hang in a very bad way, I think related to this and pytket
• [x] cirq_google.gate_set breaking change
• [x] segfault in docker when running tests. Seems to be a pytket issue. Pinning to not-the-latest-version
• [x] https://github.com/quantumlib/ReCirq/issues/214 -- not blocking. I just found it when manually testing all the pre-filled-in notebooks
• [x] molecular_data notebook issue. This is due to a change in h5py and no recent release of openfermionpyscf. Pinning h5py works for now but may cause environment issues later if other projects require a newer h5py.

• [x] ERROR: openfermion 1.0.1 has requirement cirq==0.10.0, but you'll have cirq 0.11.0 which is incompatible. https://github.com/quantumlib/OpenFermion/issues/736
• [x] If you plow ahead, it legitimately breaks because it requires the changes from https://github.com/quantumlib/OpenFermion/pull/727/ Merged May 12 but latest release is March 6.
• [x] What's the difference between openfermion 1.0.0 and 1.0.1? There's no tag for the latter. edit: there's a tag now, but:
• [ ] ~~The tag for openfermion 1.0.1 points to the wrong commit.~~
• [x] PyTket breakage with cirq.Rx https://github.com/quantumlib/ReCirq/pull/182/ the updated test in this PR is blocked by pinning to Cirq 0.11, so we can't merge yet.
• [x] PyTket updates https://github.com/quantumlib/ReCirq/pull/182/
• [x] Can no longer load in cirq.TrialResult -- https://github.com/quantumlib/Cirq/issues/4318
• [x] cirq.TrialResult loading bugfix release with https://github.com/quantumlib/Cirq/pull/4319#issuecomment-885303444
• [x] Scipy removed wrap_function https://github.com/quantumlib/ReCirq/pull/184

There's something fishy going on with the pytket notebook. When running in the doc build system it says

---------------------------------------------------------------------------
AttributeError                            Traceback (most recent call last)
<ipython-input-1-2f5bab2028d6> in <module>
     13     return pytket.device.Device({}, {}, arc)
     14 
---> 15 tk_circuit = pytket.cirq.cirq_to_tk(circuit)
     16 tk_device = _device_to_tket_device()

AttributeError: module 'pytket' has no attribute 'cirq'
AttributeError: module 'pytket' has no attribute 'cirq'

and when I try to execute it in colab it complains about a bad sympy version

ERROR: pytket 0.7.1.1 has requirement sympy~=1.6, but you'll have sympy 1.1.1 which is incompatible.

and when I run the import pytket part, the kernel dies with some information in the logs:

AttributeError: module 'sympy.core.core' has no attribute 'numbers'

Moving 3 QAOA related tutorials from the Cirq docs to ReCirq. Note that these three are just docs only: they don't use any of the additional code in ReCirq.

If it is preferred that these three files remain in the Cirq repo, then this PR would change to only add the following to the docs/_toc.yaml file:

- title: "QAOA Max-Cut"
  path: /cirq/tutorials/qaoa
- title: "QAOA: Ising Model"
  path: /cirq/tutorials/educators/qaoa_ising
- title: "Binary Paintshop"
  path: /cirq/tutorials/ionq/binary_paintshop

Since this would constitute the table of contents file for ReCirq linking out of the ReCirq repo, back to files in the Cirq repo. I thought it might be better to just move these three tutorials instead.

Towards #200.

This is blocked by openfermion and/or pytket releasing versions that support Cirq 0.12, although if you manually install Cirq 0.12 with these changes and an older pip that is ok with conflicting version specs, everything seems to work.

• Most changes are fixing deprecation warnings
• The one fix where recirq was fully broken was changing the cirq_google.gate_sets import
• Manually tested notebooks, including pre-filled-out ones

Patching CVE-2007-4559

Hi, we are security researchers from the Advanced Research Center at Trellix. We have began a campaign to patch a widespread bug named CVE-2007-4559. CVE-2007-4559 is a 15 year old bug in the Python tarfile package. By using extract() or extractall() on a tarfile object without sanitizing input, a maliciously crafted .tar file could perform a directory path traversal attack. We found at least one unsantized extractall() in your codebase and are providing a patch for you via pull request. The patch essentially checks to see if all tarfile members will be extracted safely and throws an exception otherwise. We encourage you to use this patch or your own solution to secure against CVE-2007-4559. Further technical information about the vulnerability can be found in this blog.

If you have further questions you may contact us through this projects lead researcher Kasimir Schulz.

I am running the "run-problem-generation.py" in https://github.com/quantumlib/ReCirq/blob/dcc4da9892b4e22c386e05242552660b3c558f8e/recirq/qaoa/experiments/run-problem-generation.py#L1-L60

and got the following error:

Traceback (most recent call last):
  File "c:\Users\hmy98\iCloudDrive\Desktop\Test_Generator\qaoa_tutorial.py", line 60, in <module>
    main()
  File "c:\Users\hmy98\iCloudDrive\Desktop\Test_Generator\qaoa_tutorial.py", line 52, in main
    generate_hardware_grid_problem(task)
  File "C:\Users\hmy98\iCloudDrive\Desktop\Test_Generator\venv\lib\site-packages\recirq\qaoa\experiments\problem_generation_tasks.py", line 174, in generate_hardware_grid_problem        
    problem = _get_all_hardware_grid_problems(
  File "C:\Users\hmy98\iCloudDrive\Desktop\Test_Generator\venv\lib\site-packages\recirq\qaoa\experiments\problem_generation_tasks.py", line 156, in _get_all_hardware_grid_problems       
    device_graph = _get_device_graph(device_name=device_name)
  File "C:\Users\hmy98\iCloudDrive\Desktop\Test_Generator\venv\lib\site-packages\recirq\qaoa\experiments\problem_generation_tasks.py", line 121, in _get_device_graph
    device_graph = ccr.gridqubits_to_graph_device(device.qubits)
AttributeError: 'GridDevice' object has no attribute 'qubits'

I'm trying to run the example here:

https://github.com/quantumlib/ReCirq/blob/master/docs/hfvqe/quickstart.ipynb

import numpy as np
import cirq

from recirq.hfvqe.gradient_hf import rhf_func_generator
from recirq.hfvqe.opdm_functionals import OpdmFunctional
from recirq.hfvqe.analysis import (
    compute_opdm, mcweeny_purification,
    resample_opdm, fidelity_witness,
    fidelity)
from recirq.hfvqe.third_party.higham import fixed_trace_positive_projection
from recirq.hfvqe.molecular_example import make_h6_1_3

rhf_objective, molecule, parameters, obi, tbi = make_h6_1_3()
ansatz, energy, gradient = rhf_func_generator(rhf_objective)

# settings for quantum resources
qubits = [cirq.GridQubit(0, x) for x in range(molecule.n_orbitals)]
sampler = cirq.Simulator(dtype=np.complex128)  # this can be a QuantumEngine

# OpdmFunctional contains an interface for running experiments
opdm_func = OpdmFunctional(qubits=qubits,
                           sampler=sampler,
                           constant=molecule.nuclear_repulsion,
                           one_body_integrals=obi,
                           two_body_integrals=tbi,
                           # only simulate spin-up electrons:
                           num_electrons=molecule.n_electrons // 2,
                           clean_xxyy=True,
                           purification=True
                           )


# 1.
# default to 250_000 shots for each circuit.
# 7 circuits total, printed for your viewing pleasure
# return value is a dictionary with circuit results for each permutation
measurement_data = opdm_func.calculate_data(parameters)

# 2.
opdm, var_dict = compute_opdm(measurement_data, return_variance=True)
opdm_pure = mcweeny_purification(opdm)

# 3.
raw_energies = []
raw_fidelity_witness = []
purified_eneriges = []
purified_fidelity_witness = []
purified_fidelity = []
true_unitary = ansatz(parameters)
nocc = molecule.n_electrons // 2
nvirt = molecule.n_orbitals - nocc
initial_fock_state = [1] * nocc + [0] * nvirt

# 1000 repetitions of the measurement
for _ in range(1000):  
    new_opdm = resample_opdm(opdm, var_dict)
    raw_energies.append(opdm_func.energy_from_opdm(new_opdm))
    raw_fidelity_witness.append(
        fidelity_witness(target_unitary=true_unitary,
                         omega=initial_fock_state,
                         measured_opdm=new_opdm)
    )
    # fix positivity and trace of sampled 1-RDM if strictly outside
    # feasible set
    w, v = np.linalg.eigh(new_opdm)
    if len(np.where(w < 0)[0]) > 0:
        new_opdm = fixed_trace_positive_projection(new_opdm, nocc)

    new_opdm_pure = mcweeny_purification(new_opdm)
    purified_eneriges.append(opdm_func.energy_from_opdm(new_opdm_pure))
    purified_fidelity_witness.append(
        fidelity_witness(target_unitary=true_unitary,
                         omega=initial_fock_state,
                         measured_opdm=new_opdm_pure)
    )
    purified_fidelity.append(
        fidelity(target_unitary=true_unitary,
                 measured_opdm=new_opdm_pure)
    )
print("Canonical Hartree-Fock energy ", molecule.hf_energy)
print("True energy ", energy(parameters))
print("Raw energy ", opdm_func.energy_from_opdm(opdm),
      "+- ", np.std(raw_energies))
print("Raw fidelity witness ", np.mean(raw_fidelity_witness).real,
      "+- ", np.std(raw_fidelity_witness))
print("purified energy ", opdm_func.energy_from_opdm(opdm_pure),
      "+- ", np.std(purified_eneriges))
print("Purified fidelity witness ", np.mean(purified_fidelity_witness).real,
      "+- ", np.std(purified_fidelity_witness))
print("Purified fidelity ", np.mean(purified_fidelity).real,
      "+- ", np.std(purified_fidelity))

---------------------------------------------------------------------------
KeyError                                  Traceback (most recent call last)
File ~/.conda/envs/qiskit/lib/python3.10/site-packages/pandas/core/indexes/base.py:3621, in Index.get_loc(self, key, method, tolerance)
   3620 try:
-> 3621     return self._engine.get_loc(casted_key)
   3622 except KeyError as err:

File ~/.conda/envs/qiskit/lib/python3.10/site-packages/pandas/_libs/index.pyx:136, in pandas._libs.index.IndexEngine.get_loc()

File ~/.conda/envs/qiskit/lib/python3.10/site-packages/pandas/_libs/index.pyx:163, in pandas._libs.index.IndexEngine.get_loc()

File pandas/_libs/hashtable_class_helper.pxi:5198, in pandas._libs.hashtable.PyObjectHashTable.get_item()

File pandas/_libs/hashtable_class_helper.pxi:5206, in pandas._libs.hashtable.PyObjectHashTable.get_item()

KeyError: '(0, 1)'

The above exception was the direct cause of the following exception:

KeyError                                  Traceback (most recent call last)
Input In [6], in <cell line: 8>()
      5 measurement_data = opdm_func.calculate_data(parameters)
      7 # 2.
----> 8 opdm, var_dict = compute_opdm(measurement_data, return_variance=True)
      9 opdm_pure = mcweeny_purification(opdm)
     11 # 3.

File ~/.conda/envs/qiskit/lib/python3.10/site-packages/recirq/hfvqe/analysis.py:66, in compute_opdm(results_dict, return_variance)
     64     q0, q1 = qubits[pair_idx:pair_idx + 2]
     65     qA, qB = permutation[pair_idx:pair_idx + 2]
---> 66     opdm[qA, qB] += np.mean(data[q1] - data[q0], axis=0) * 0.5
     67     opdm[qB, qA] += np.mean(data[q1] - data[q0], axis=0) * 0.5
     69 if return_variance:
     70     # get covariance matrices

File ~/.conda/envs/qiskit/lib/python3.10/site-packages/pandas/core/frame.py:3505, in DataFrame.__getitem__(self, key)
   3503 if self.columns.nlevels > 1:
   3504     return self._getitem_multilevel(key)
-> 3505 indexer = self.columns.get_loc(key)
   3506 if is_integer(indexer):
   3507     indexer = [indexer]

File ~/.conda/envs/qiskit/lib/python3.10/site-packages/pandas/core/indexes/base.py:3623, in Index.get_loc(self, key, method, tolerance)
   3621     return self._engine.get_loc(casted_key)
   3622 except KeyError as err:
-> 3623     raise KeyError(key) from err
   3624 except TypeError:
   3625     # If we have a listlike key, _check_indexing_error will raise
   3626     #  InvalidIndexError. Otherwise we fall through and re-raise
   3627     #  the TypeError.
   3628     self._check_indexing_error(key)

KeyError: '(0, 1)'

seems like this might be a cirq/recirq version issue? Here are my versions

cirq-core                     0.15.0
cirq-google                   0.15.0

recirq                        0.1.dev0

• include job timestamp
• re-think aggregation strategy

Background: the analysis code extracts the results into a pandas dataframe. You need to aggregate the quantities to make plots and create fits.

Previously: I knew additional columns (i.e. fields) would be added over time to the results dataframe. I thought I could be clever with the aggregation to make the addition of new fields automatic. This was done with the groupby_all_except function which inverts the api of groupby.

Then: we actually need a custom aggregation for the job_finished_time field. Every time the dataframe gets aggregated down, I've chosen to pick the "last" time of the group. You can imagine choosing the first or average. In any event: the groupby_all_except approach didn't make this magically work. One still needs to plumb through the new field with care.

This PR: make our groupby columns and the aggregation functions / output columns explicit everywhere. Going forward, this shouldn't behave badly if a new field is added. Although if you want to actually have the new field show up after using the aggregation/fitting functions you have to add it to the [...]_y_cols mapping in each relevant function.

• SKModelQAOASpec and sk_model_qaoa_spec_to_exe wraps existing QAOA code into the cirq_google.workflow tools so this can serve as an algorithmic benchmark.
• scripts and recirq.algorithmic_benchmark_library entry to follow
• Small changes to the circuit generation code to support generating lightly-compiled executables so the benchmark can target multiple gatesets following https://github.com/quantumlib/Cirq/pull/5336