pystaq

pystaq is a Python 3 wrapper for the staq tool suite.

Installation

pystaq can be installed using pip:

pip install git+https://github.com/softwareQinc/staq

Overview

To parse a circuit, use the function pystaq.parse_file, which takes a file path as input, or pystaq.parse_str, which accepts an OpenQASM 2.0 program string.

The library provides the following tools:

desugar
inline
synthesize_oracles
simplify
rotation_fold
map
get_resources
output_cirq
output_projectq
output_qsharp
output_quil

Each function takes as input a parsed program, followed by any options supported by the corresponding staq tool.

---

Example:

>>> import pystaq

>>> p = pystaq.parse_str('''
OPENQASM 2.0;
include "qelib1.inc";

qreg q[2];
gate cphase(theta) a,b {
  rz(theta/2) a;
  rz(theta/2) b;
  cx a,b;
  rz(-theta/2) b;
  cx a,b;
}
cphase(pi/4) q[0],q[1];
''')

>>> pystaq.inline(p, clear_decls=True)

>>> p
OPENQASM 2.0;
include "qelib1.inc";

qreg q[2];
rz((pi/4)/2) q[0];
rz((pi/4)/2) q[1];
cx q[0],q[1];
rz(-(pi/4)/2) q[1];
cx q[0],q[1];

Device generator

The pystaq.Device class can be used to create custom devices for mapping. It has the methods add_edge and set_fidelity.

---

Example:

import pystaq
dev = pystaq.Device(4)           # 4 qubits
dev.add_edge(0, 1)               # default is undirected, with fidelity = FIDELITY_1 from staq
dev.add_edge(2, 3, fidelity=0.8, directed=True)
dev.set_fidelity(0, 0.9)         # single qubit fidelity
with open('device.json', 'w') as f:
    f.write(str(dev))

This produces the following device.json file, which can then be used by pystaq.map:

{
  "couplings": [
    {
      "control": 0,
      "target": 1
    },
    {
      "control": 1,
      "target": 0
    },
    {
      "control": 2,
      "fidelity": 0.8,
      "target": 3
    }
  ],
  "name": "Custom Device",
  "qubits": [
    {
      "fidelity": 0.9,
      "id": 0
    },
    {
      "id": 1
    },
    {
      "id": 2
    },
    {
      "id": 3
    }
  ]
}

Custom Bindings

pystaq was created using pybind11. See pystaq/staq_wrapper.cpp for many examples of how to wrap a circuit transformation.

For more details, see also our Quantum++ wrapper pyqpp.