README.md

Taichi-Q

Introduction

Taichi-Q is an open source framework for quantum computation simulation based on Taichi that support both CPU and GPU.

Taichi-Q is originally a 2nd prized Taichi Hackathon Project, detailed introduction (in Chinese) is available here.

This package is in a early release version. We're ready to update more features in the future.

Intallation

• From PyPI

pip install taichi-q

Usage

Import the Packages

# Required
from taichi_q import Engine, Gate

# Optional
import numpy as np
import taichi as ti

Setup Simulator Engine

Warning!! Simulator Engine could only be initialized once.

For CPU

eng=Engine(num_qubits=3,state_init=0,device='cpu')

For GPU

eng=Engine(num_qubits=3,state_init=0,device='gpu')

Initialize Qubits

• All $|0\rangle$ or $|1\rangle$

eng=Engine(num_qubits=3, state_init=0, device='cpu')
eng=Engine(num_qubits=3, state_init=1, device='cpu')

• Choose $|0\rangle$ or $|1\rangle$ of each qubit with list/tuple/ndarray

eng=Engine(num_qubits=3, state_init=[0,1,0], device='gpu')
eng=Engine(num_qubits=3, state_init=(0,1,0), device='gpu')
eng=Engine(num_qubits=3, state_init=np.array([0,1,0]), device='gpu')

• Set complex qubit state with np.array(dtype=complex)

eng = Engine(
    num_qubits=3,
    state_init=[[-1/np.sqrt(2), j/np.sqrt(2)], [1, 0], [0, 1]],
    device='cpu')

Quantum Gate Operators

Quantum Gates could be found in taichi_q.gates. Support $H(), X(), Y(), Z(), S(), T(), swap()$ $U(\theta, \phi, \lambda), R_x(\theta), R_y(\theta), R_z(\theta)$ $QFT(n), iQFT(n)$, and all controlled gates.

• Apply single-qubit gate to target qubit (e.g. H)

eng.Ops(Gate.H(), [0])

• Apply muti-qubit gate to target qubits (e.g. swap)

eng.Ops(Gate.swap(), [0,1])

• Apply controlled-qubit gate to target qubits (e.g. CNOT=CX)

eng.Ops(Gate.X(), [0], [1])

• If you want to print Operated Gate, Tgt and Ctl on the terminal

eng.Ops(Gate.QFT(4), [0,1,2,3], [4], print_output=True)

# Output:
# OPS: QFT  Tgt: [0,1,2,3]   Ctl [4]

Measure the result of a qubit

Notice! Measure is an irreversible process. State of the measured qubit would collapsed into $|0\rangle$ or $|1\rangle$

q0_result=eng.Measure(0)

Check the state of all qubits without measuring any qubit

Notice! This is a cheating method from simulator. It's not available for real quantum computer.

Check the state of all qubits is useful for quantum computation algorithm design and debug.

• Print all qubit states

eng.State_Check(print_state=True)

# Output:
# Q: (0, 0, 0)   State:[+0.0000+0.0000j]   P:0.0000
# Q: (0, 0, 1)   State:[+0.0000+0.0000j]   P:0.0000
# Q: (0, 1, 0)   State:[+0.0000+0.0000j]   P:0.0000
# Q: (0, 1, 1)   State:[+0.0000+0.0000j]   P:0.0000
# Q: (1, 0, 0)   State:[+0.0000+0.0000j]   P:0.0000
# Q: (1, 0, 1)   State:[+0.0000+0.0000j]   P:0.0000
# Q: (1, 1, 0)   State:[-0.4472+0.0000j]   P:0.2000
# Q: (1, 1, 1)   State:[+0.0000+0.8944j]   P:0.8000

• Display Box-plot of qubit state probability

eng.State_Check(plot_state=True)

Print Quantum Circuit

Quantum Gates and its order are recorded by the engine, and could be displayed on the terminal.

eng.circuit_print()

# Output:
# Q0 →|' ' ' ' '■' 'H' 'M' ' ' ' ' '■'|→ Q0
# Q1 →|'H' '■' 'X' ' ' ' ' 'M' '■' ' '|→ Q1
# Q2 →|' ' 'X' ' ' ' ' ' ' ' ' 'X' 'Z'|→ Q2

Visualize Quantum Circuit (with ti.GUI)

The circuit visualization provides a more elegant approach for circuit visualization, based on ti.GUI.

eng.circuit_visualize()

Examples

• Generate Bell-State
• Quantum Teleportation
• Grover Searching Algorithm

Development

Welcome any contribution!

License

Apach-2.0