Sampling

There’re 2 ways to do sampling after simulation.

Note

Sampling implementation is preliminary in 0.2.x. APIs are subject to change.

Sampling by iterating simulations

Simulator.sample(circuit, ref_array, n_samples) method runs simulations for n_samples times, and returns measurement results as ObservationList.

circuit : a circuit to execute sampling ref_array : a list of references to be sampled n_samples : number of sampling shots

Bit ordering of returned ObservationList is defined by ref_array.

qregs = new_qregs(n_qregs)
refs = new_references(2)

ops = [ ... gate sequence ...,
        measure(refs[0], qregs[n],
        ... gate sequence ...,
        measure(refs[1], qregs[m],
        ... gate sequence ...
      ]

sim = qgate.simulator.cpu()
obslist = sim.sample(ops, refs, 100) # sampling results are returned as observation list.
hist = obslist.histgram()   # getting histgram from observations.

Sampling pool

Sampling pool is for efficient and fast sampling by using pre-calculated probability vector.

In Qgate 0.2.x, sampling implementation has limitations shown below:

1. Quantum circuits is not allowed to have measurement,

   Measurement oprations destory quantum coherence, so should not be included.

2. Quantum circuits is not allowed to have if-clause.

   if-clause uses measurement results to branch execution of quantum circuits.

Sampling pool is created by calling Qubits.create_sampling_pool(qregs). The parameter, qregs, is a list of qregs to be sampled.

By calling SamplingPool.sample(n_samples), sampling results are returned as ObservationList.

Bit ordering of returned ObservationList is defined by the qregs parameter in Qubits.create_sampling_pool(qregs).

ops = ... preparing circuit ...

sim = qgate.simulator.cpu()
sim.run(ops)

qregs = ... list of quantum registers which will be sampled ...
sampling_pool = sim.qubits.create_sampling_pool(qregs)

obslist = sampling_pool.sample(n_samples) # sampling results are returned as observation list.
hist = obslist.histgram()                 # getting histgram from observations.