PicoQuant.jl

Author JuliaQX
Popularity
10 Stars
Updated Last
2 Years Ago
Started In
June 2020

PicoQuant

Build Status

PicoQuant is a quantum circuit simulation framework being developed as part of the QuantEx project. This is a PRACE funded project to develop quantum circuit simulation tools capable of running on the classical exa-scale compute clusters expected to be deployed in the coming years. This package is a prototype to test out ideas and approaches that we hope to improve on in subsequent iterations.

The best way to get started is to:

  • Install the prototype locally by following the instructions below
  • Follow the series of tutorial notebooks in the notebooks folder
  • For more details on function interfaces read the online docs
  • Get involved by posting an issue or submitting a pull request

Install and example

PicoQuant can be installed from the Julia REPL with the following commands

import Pkg
Pkg.add("PicoQuant)

Tests can be run with

import Pkg
Pkg.test()

A simple script to use PicoQuant to simulate a 3 qubit GHZ circuit consists of the following

using PicoQuant

tn = TensorNetworkCircuit(3)
add_gate!(tn, gate_tensor(:H), [1])
add_gate!(tn, gate_tensor(:CX), [1, 2])
add_gate!(tn, gate_tensor(:CX), [2, 3])
add_input!(tn, "000")
full_wavefunction_contraction!(tn, "vector")
output = load_tensor_data(tn, :result)

Detailed installation instructions

The version of PicoQuant that the above methods installs is the version that is registered with the Julia repository. These instructions show how to clone from the repository and install this version.

  1. If Julia is not already installed, download and install Julia from the Julialang website
  2. Clone this repository using git clone https://github.com/ICHEC/PicoQuant.jl.git
  3. Start the Julia REPL and navigate to the PicoQuant folder, activate and instantiate the environment and then build PicoQuant.
import Pkg
Pkg.activate(".")
Pkg.instantiate()
Pkg.build("PicoQuant")

This should install dependencies specified in the Project.toml file and carry out any package specific build tasks (detailed in deps/build.jl file). Currently PicoQuant uses some functionality from qiskit. This is installed in the python environment used by PyCall during the build. See below for details about using different python environments.

Running the unittests

Unittests can be run from the PicoQuant root folder with

julia --project=. test/runtests.jl

This will run all the unittests. It's possible to run a subset of the unittests by passing the name of the testset. For example to run the layer3 tests contained in the test/layer3_tests.jl script one would run

julia --project=. test/runtests.jl layer3_tests

Running standalone scripts

Standalone executable scripts are located in the bin/ folder. As an example we show how to run the qasm2tng.jl script to convert the qft_3.qasm file to json format.

julia --project=. bin/qasm2tng.jl --qasm qft_3.qasm --output qft_3.json

Starting a notebook server

Much of the prototyping and development is done in jupyter notebooks which provides instant feedback and speeds development. To start a jupyer notebook from the Julia REPL, enter

using IJulia
notebook()

This should open a browser window showing the home folder.

Using different python environments

Note that PicoQuant makes use of python libraries via the PyCall.jl package. On Linux systems this will use the python3 binary in the path (or python if there is no python3 binary found). On windows and macOS systems it will create a dedicated conda environment which will reside at ${HOME}/.julia/conda/3. The required python packages are installed as part of the build of PicoQuant. To use a different python environment, the PYTHON environment variable must be set to point to the python binary and PyCall needs to be (re)built. For example to create a new conda environment at ~/.julia/conda/picoquant_env, one would follow the steps

Use conda to create the environment from the command line

conda -p ~/.julia/conda/picoquant_env python=3.7

Then from the Julia REPL

ENV["PYTHON"] = "~/.julia/conda/picoquant_env"
]build PyCall

Building the documentation

The package uses Documenter.jl to generate html documentation from the sources. To build the documentation, run the make.jl script from the docs folder.

julia --project=docs docs/make.jl

The documentation will be placed in the build folder and can be hosted locally by starting a local http server with

cd docs/build && python3 -m http.server

As part of the CI this documentation is automatically built and hosted via github pages at https://ICHEC.github.io/PicoQuant.jl/dev