GUIDELINE.md

Procedure for quantum state preparation (QSP) experiment.

QSP is presented clearly in this publication. The basic idea is that we have two unitary $U(\theta)$ and $V^{\dagger}$, we optimize the parameter $\theta$ until $\theta^{}$ such that $U(\theta^{})V^{\dagger}=I$.

So we will do experiment with various $U$, $V^{\dagger}$ and optimize strategy.

Some notation for hyperparameter:

(important)

• Ansatz: $U$ (defined in qsee.ansatz)
• Optimize circuit: $u$ (ansatz when repeat $L$ times)
• State: $V^{\dagger}$ (defined in qsee.core.state)
• Number of qubit: $n$ (num_qubits)
• Number of layer (for ansatz): $L$ (num_layers)
• $\theta$: theta (if $\theta$ is scalar), thetas (if $\theta$ is 1-d numpy array) and thetass (if $\theta$ is 2-d numpy array).

(not important)

• Number of iteration: $n_{iter}$ (num_iter / iter)
• Optimizer: sgd, adam or qng family.

Save result as qspobj follow this format name, directly in qsp folder:

[state]_[ansatz]_[num_qubits]_[num_layers].qspobj