Tangelo v0.3.4 is here

[ValentinS4t1qbit]

Howdy’ Tangelites ! v0.3.4 is now available.

If you think this project is a good idea, please click right now on the ⭐ button on the top-right corner of your screen, and consider downloading Tangelo. It helps us make this project more visible and to understand how relevant it is to the community. Seriously, it’s small but it means a bunch to us right now :)

Credits

Cheers to @JamesB-1qbit, @AlexandreF-1qbit, @KrzysztofB-1qbit and @ValentinS4t1qbit for contributions to the codebase, and to @chrislzhao for his feature request and feedback, integrated in time for this release. Bug reports and feature suggestions are super useful: this is what helps our community make tools that really help people. And that’s what this project is all about.

Highlights of this release:

Python version bump to 3.8+

Python 3.7 is no longer maintained by the Python team: Tangelo now requires you to have Python 3.8 or more in your environment. The latest version of Python is currently 3.11; if you happen to have issues with any of the newer versions please let us know. We’ll try to extend our testing and make things work for you.

You don't have to choose between quantum frameworks. Use all of them.

It’s now really easy to convert existing projects between Tangelo and your favorite frameworks; now both quantum circuits and qubit operators can be converted in a single line of code, in either direction. There are so many platforms with unique features and cool backends to explore: we think that freedom is essential to the research in our field. Two simple functions allow you to convert your quantum circuits and qubit operators between a variety of popular formats, including the recently-added Pennylane.

You have an existing project in Qiskit, want to use Tangelo to explore some workflows and then take this to a quantum device available on Amazon Braket ? Go for it. Neither the source or target format need to be "tangelo", but the conversion uses it as an intermediary format under the hood.

from tangelo.linq import translate_circuit, translate_operator

qiskit_circ = translate_circuit(tangelo_circ, target="qiskit")
tangelo_op = translate_operator(qiskit_op, source="qiskit", target="tangelo")

If you can’t find the format you need, we’d love you to submit a feature request, or even better: try to contribute it to the codebase and let your peers benefit from it. We’re here to support you with the process, may you be a seasoned developer or not.

Mid-circuit measurements

Additional features are now available for simulating circuits with mid-circuit measurements. The simulate method can now save mid-circuit measurements. You can also explore the statevector resulting from a set of mid-circuit measurements, or retrieve the deterministic state corresponding to a desired mid-circuit measurement result. These features are implemented as optional arguments to simulate (see small example below).

Many algorithms depend on a certain measurement result to signify a successful quantum circuit run. These algorithms include Linear Combination of Unitaries (LCU), Quantum Signal Processing (QSP), symmetry projection, Rodeo Algorithm, probabilistic imaginary time evolution, quantum Lanczos, and more ! It is now possible to link these algorithms with further quantum operations intrinsically in Tangelo.

c = Circuit([Gate("H", 0), Gate("MEASURE", 0), Gate("CNOT", 1, 0)])
sim = get_backend()
sim.simulate(c, desired_meas_result="0")[0])      # Output:  {'00': 1.0}
sim.simulate(c, desired_meas_result="1")[0])      # Output:  {'11': 1.0}     

UHF mean-field reference: towards better modeling of open-shell systems

The unrestricted Hartree-Fock mean field is now supported in Tangelo, during the creation of the SecondQuantizedMolecule. This mean-field method gives a better starting state for the search of the ground state for open-shell chemical systems, i.e. molecules with different numbers of electrons of each spin. A given algorithm would then give a better representation of the state, thus improving the chemical properties computed from this numerical representation.

Those systems are important to study chemical kinetics, where elementary reactions are primordial. In other words, there is a high probability that high-energy intermediate molecules could either be charged or a radical: UHF is generally a better mean-field description of the latter. The Tangelo team is currently working towards supporting UHF mean-field with some variational algorithms and problem decomposition techniques, and using it for research: more to come later this year.

Algorithms

• You can now create time-evolution circuits for time-dependant Hamiltonians efficiently using the discrete clock construction (https://arxiv.org/abs/2203.11353) and the multiproduct trotter evolution (https://arxiv.org/abs/1907.11679). added multi-product, grid_circuits and discrete_clock #257

• Grid-like basis functions are hypothesized to be the most efficient for fault-tolerant quantum computers. Tangelo now has functions that generate the time-evolution for the kinetic energy operator and potentials of the form a + bx + cx^2. added multi-product, grid_circuits and discrete_clock #257

• The pUCCD ansatz is now available for VQE-based solvers. This chemistry-inspired ansatz implements the hard-core boson Hamiltonian approximation and thus requires only 3 measurement bases, at the cost of a potentially reduced energy / state accuracy (https://arxiv.org/abs/2002.00035). pUCCD ansatz #251

Tutorials

There is now a dedicated Github repository, that you can explore directly through Github or Github pages. This space welcomes tutorials and scripts from all, and is an opportunity to showcase your Tangelo projects: we’re happy to list projects involving Tangelo there as well. Don’t be shy :)

We tried to sort things to help you find what’s relevant to you, and added a few new things there:

• Alexandre Fleury came across a recent publication about Graph |Q><C| and wanted to put Tangelo to the test. With a simple script (~60 lines of actual code), he drew what he needed from the toolboxes in Tangelo and reproduced some of the results - he even pushed a bit further with very little effort, by looking at different ansatze -. A pleasant experience suggesting that Tangelo can help people keep up with state-of-the-art and accelerate research.

• James Brown uploaded a notebook showing how he easily integrated the Clifford simulator from Cirq as a custom backend (not supported as a built-in in Tangelo) to his Tangelo workflows. This integration resulted in research that led to the manuscript posted on arXiv titled Iterative Qubit Coupled Cluster using only Clifford circuits.

• Check out the notebook behind the demo we realized for the IBM Quantum Summit in October 2022. It shows how Tangelo seamlessly integrates with IBM Quantum, on a simple use case leveraging the error-mitigation techniques of both Tangelo and IBM’s technology. You can just zoom to the last section if what’s relevant to you is how the qiskit-runtime integration works, the gist of it is very simple and just 3-4 lines of code.

---

This discussion was created from the release Tangelo v0.3.4 is here.