Tutorial: Running an Experiment

See also

The complete source code of this tutorial can be found in

Running an Experiment.ipynb

The example dataset can be downloaded here.

This notebook presents a structure for setting up experiments using a combination of quantify-scheduler and quantify-core. quantify-scheduler provides a high-level interface with the hardware, allowing users to abstract hardware-specific nuances. quantify-core, on the other hand, serves as an experiment management tool, using quantify-scheduler as its hardware interface. This allows users to manage, execute, and analyze experiments easily.

The following is a general workflow for using Quantify

• 1. Initial Setup

  • Set the directory for data storage for the experiment

  • Initialize the MeasurementControl and InstrumentCoordinator objects

• 2. Device Setup

  • Set up a device compilation configuration for the device under test

• 3. Hardware Setup

  • Connect to the control hardware

  • Set up the hardware compilation configuration

• 4. Schedule Definition and Compilation

  • Create a schedule containing the timeline of operations for the experiment

  • Compile the schedule to control-hardware code

  • Visualize the schedule

• 5. Experiment Execution and Analysis

  • Setup MeasurementControl to run the experiment

  • Run the experiment

  • Analyze the results

1. Initial Setup

We first set up the directory in which all experimental data will be stored and managed.

from quantify_core.data import handling as dh
dh.set_datadir()

Data will be saved in:
/root/quantify-data

Next, we need to initialize two classes: MeasurementControl for managing the experiment and InstrumentCoordinator for managing the control hardware.

from quantify_core.measurement.control import MeasurementControl
from quantify_scheduler.instrument_coordinator import InstrumentCoordinator

measurement_control = MeasurementControl("measurement_control")
instrument_coordinator = InstrumentCoordinator("instrument_coordinator")

2. Device Setup

We set up the quantum device on which we perform the actual experiments. This one-qubit chip is represented by QuantumDevice where we add a single transmon qubit (represented by BasicTransmonElement) q0 to it.

# Device parameters
ACQ_DELAY = 100e-9
FREQ_01 = 4e9
READOUT_AMP = 0.1
READOUT_FREQ = 4.3e9
PI_PULSE_AMP = 0.15
LO_FREQ_QUBIT = 3.9e9
LO_FREQ_READOUT = 4.5e9

from quantify_scheduler.device_under_test.quantum_device import QuantumDevice
from quantify_scheduler.device_under_test.transmon_element import BasicTransmonElement

single_qubit_device = QuantumDevice("single_qubit_device")
single_qubit_device.instr_instrument_coordinator(instrument_coordinator.name)

q0 = BasicTransmonElement("q0")
single_qubit_device.add_element(q0)

# Assign device parameters to transmon element
q0.measure.pulse_amp(READOUT_AMP)
q0.clock_freqs.readout(READOUT_FREQ)
q0.clock_freqs.f01(FREQ_01)
q0.measure.acq_delay(ACQ_DELAY)
q0.rxy.amp180(PI_PULSE_AMP)

Quantum Devices and Elements

More information on quantum devices and elements can be found in Tutorial: Operations and Qubits.

3. Hardware Setup

Let us now set up the connections to the control hardware. In this example, we use a dummy Qblox device that is created via an instance of the Cluster class, and is initialized with a dummy configuration consisting of a readout module in slot 1 and a control module in slot 2.

from qblox_instruments import Cluster, ClusterType
from quantify_scheduler.instrument_coordinator.components.qblox import ClusterComponent
cluster = Cluster(
    "cluster",
    dummy_cfg={
        1: ClusterType.CLUSTER_QRM_RF,
        2: ClusterType.CLUSTER_QCM_RF,
    },
)

ic_cluster = ClusterComponent(cluster)
instrument_coordinator.add_component(ic_cluster)

The last part of setting up the hardware is to define the HardwareCompilationConfig and attach it to our single_qubit_device. The hardware compilation configuration is a pydantic datastructure, parsed either from a file or from a Python dictionary. It contains all of the information about the instruments used to run the experiment and is used to compile the schedule to hardware. For more information on this datastructure, please refer to the explanation in the User Guide.

hardware_comp_cfg = {
    "config_type": "quantify_scheduler.backends.qblox_backend.QbloxHardwareCompilationConfig",
    "hardware_description": {
        f"{cluster.name}": {
            "instrument_type": "Cluster",
            "ref": "internal",
            "modules": {
                "1": {
                    "instrument_type": "QRM_RF"
                },
                "2": {
                    "instrument_type": "QCM_RF"
                },
            },
        },
    },
    "hardware_options": {
        "modulation_frequencies": {
            "q0:res-q0.ro": {"lo_freq": LO_FREQ_READOUT},
            "q0:mw-q0.01": {"lo_freq": LO_FREQ_QUBIT},
        },
    },
    "connectivity": {
        "graph": [
            (f"{cluster.name}.{cluster.module1.name.split('_')[-1]}.complex_output_0", "q0:res"),
            (f"{cluster.name}.{cluster.module2.name.split('_')[-1]}.complex_output_0", "q0:mw")
        ]
    },
}

# Tie hardware config to device
single_qubit_device.hardware_config(hardware_comp_cfg)

4. Schedule Definition and Compilation

Now we must create a schedule, where we define the set of operations that we wish to perform on the device under test. We define the schedule independent of the hardware configuration and rely on Quantify’s ability to compile a schedule to hardware for converting it to hardware-level commands. For this tutorial, we will define a simple schedule that will run a T1 experiment to determine the relaxation time of our qubit. For various delay times tau, we repeatedly excite the qubit, wait tau seconds and then measure the qubit.

from quantify_scheduler import Schedule
from quantify_scheduler.operations.gate_library import Measure, Reset, X

def t1_sched(times, repetitions=1):
    schedule = Schedule("T1", repetitions)
    for i, tau in enumerate(times):
        schedule.add(Reset("q0"), label=f"Reset {i}")
        schedule.add(X("q0"), label=f"pi {i}")
        schedule.add(
            Measure("q0", acq_index=i),
            ref_pt="start",
            rel_time=tau,
            label=f"Measurement {i}",
        )
    return schedule

Pre-defined Schedules

The T1 schedule can be imported with from quantify_scheduler.schedules import t1_sched.

For more pre-defined schedules, see quantify_scheduler.schedules.

We can inspect the details of the schedule in three different ways: via (1) circuit and (2) pulse diagrams, or through (3) a timing table containing the timing and other information of each operation.

The plot_circuit_diagram() method can directly be used to display the quantum circuit that corresponds to the schedule that we defined above.

t1_schedule = t1_sched(times=[1e-6])
t1_schedule.plot_circuit_diagram();

For displaying the pulse diagram and timing table, the schedule first needs to get timing information for each operation in the schedule. This is achieved by compiling the schedule using SerialCompiler. Please note that below in section 5. Experiment Execution and Analysis, the compilation is handled by MeasurementControl internally; it is done here only to enable displaying the timing table and pulse diagram.

Compiling to Hardware

More information on compilation can be found in Tutorial: Compiling to Hardware.

from quantify_scheduler.backends.graph_compilation import SerialCompiler

compiler = SerialCompiler(name="compiler", quantum_device=single_qubit_device)
compiled_schedule = compiler.compile(schedule=t1_schedule)

Each operation is compiled into pulses that can be viewed and inspected via the plot_pulse_diagram() method.

compiled_schedule.plot_pulse_diagram(plot_backend="plotly")

By passing the plotly backend, we made the above diagram interactive.

Lastly, the timing table can simply be accessed via the timing_table property of a schedule.

compiled_schedule.timing_table

	waveform_op_id	port	clock	abs_time	duration	is_acquisition	operation	wf_idx	operation_hash
0	Reset('q0')_acq_0	None	cl0.baseband	0.0 ns	200,000.0 ns	False	Reset('q0')	0	-7175969483310682096
1	X(qubit='q0')_acq_0	q0:mw	q0.01	200,000.0 ns	20.0 ns	False	X(qubit='q0')	0	-4892235067269535681
2	ResetClockPhase(clock='q0.ro',t0=0)_acq_0	None	q0.ro	201,000.0 ns	0.0 ns	False	ResetClockPhase(clock='q0.ro',t0=0)	0	-2001240957906938601
3	SquarePulse(amp=0.1,duration=3e-07,port='q0:res',clock='q0.ro',reference_magnitude=None,t0=0)_acq_0	q0:res	q0.ro	201,000.0 ns	300.0 ns	False	SquarePulse(amp=0.1,duration=3e-07,port='q0:res',clock='q0.ro',reference_magnitude=None,t0=0)	0	-1660728598727302108
4	SSBIntegrationComplex(port='q0:res',clock='q0.ro',duration=1e-06,acq_channel=0,acq_index=0,bin_mode='average',phase=0,t0=1e-07)_acq_0	q0:res	q0.ro	201,100.0 ns	1,000.0 ns	True	SSBIntegrationComplex(port='q0:res',clock='q0.ro',duration=1e-06,acq_channel=0,acq_index=0,bin_mode='average',phase=0,t0=1e-07)	0	-4695369217779849369

Device-layer Schedules (Pulses)

More information on defining schedules using pulses can be found in Tutorial: Schedules and Pulses.

Circuit-layer Schedules (Gates and Measurements)

More information on defining schedules using circuit-layer operations and mixing with pulses can be found in Tutorial: Operations and Qubits.

5. Experiment Execution and Analysis

We can now configure our MeasurementControl to run the experiment. In this case, we will perform a 1-dimensional sweep using a qcodes.instrument.parameter.ManualParameter. Sweeping such a parameter will physically change the hardware output as the sweep is performed.

import numpy as np
from qcodes.instrument.parameter import ManualParameter
from quantify_scheduler.gettables import ScheduleGettable

# Configure the settable
time = ManualParameter("sample", label="Sample time", unit="s")
time.batched = True

times = np.linspace(start=1.6e-7, stop=4.976e-5, num=125)

# Configure the gettable
gettable = ScheduleGettable(
    quantum_device=single_qubit_device,
    schedule_function=t1_sched,
    schedule_kwargs={"times": times},
    batched=True
)

# Configure MeasurementControl
measurement_control.settables(time)
measurement_control.setpoints(times)
measurement_control.gettables(gettable)

In the above example, we use settables() and gettables() together with batched=True to instruct measurement_control how to execute the experiment. Instead of iteratively processing each value in the times array, all are processed in one batch.

Configuring MeasurementControl

More information on configuring MeasurementControl can be found in the user guide of quantify-core.

Once MeasurementControl is all set up, we can run the experiment.

dataset = measurement_control.run()

The result of the experiment is assigned to the variable dataset, but also saved to disk (as set by set_datadir() at the start of this tutorial).

To analyze the experiment results, we can choose from the variety of classes from the quantify_core.analysis package. For a T1 experiment, we use the T1Analysis class which is used to fit the data, extract relevant parameters and visualize the result. In this tutorial, we run the experiment on a dummy cluster, and therefore the dataset that we obtain is empty. For demonstrational purposes, we show how it would look on an actual dataset. This example dataset can be downloaded here.

import xarray

dataset = xarray.open_dataset("../examples/dataset.hdf5")

from quantify_core.analysis import T1Analysis

T1Analysis(dataset=dataset).run().display_figs_mpl()

Analyzing Datasets

More information on analyzing datasets can be found in the user guide of quantify-core.