# Repository: https://gitlab.com/quantify-os/quantify-scheduler
# Licensed according to the LICENCE file on the main branch
"""Module containing Qblox InstrumentCoordinator Components."""
from __future__ import annotations
import copy
import logging
import os
import re
import warnings
from abc import ABC, abstractmethod
from dataclasses import dataclass
from functools import partial
from math import isnan
from typing import (
TYPE_CHECKING,
Any,
Callable,
Hashable,
Union,
)
from uuid import uuid4
import numpy as np
from qblox_instruments import (
Cluster,
ConfigurationManager,
SequencerStates,
SequencerStatus,
)
from xarray import DataArray, Dataset
from quantify_core.data.handling import get_datadir
from quantify_scheduler.backends.qblox import constants, driver_version_check
from quantify_scheduler.backends.qblox.enums import (
ChannelMode,
LoCalEnum,
SidebandCalEnum,
TimetagTraceType,
)
from quantify_scheduler.backends.qblox.helpers import (
single_scope_mode_acquisition_raise,
)
from quantify_scheduler.backends.qblox.operation_handling.bin_mode_compat import (
QRM_COMPATIBLE_BIN_MODES,
QTM_COMPATIBLE_BIN_MODES,
IncompatibleBinModeError,
)
from quantify_scheduler.backends.types.qblox import (
AnalogModuleSettings,
AnalogSequencerSettings,
BasebandModuleSettings,
BaseModuleSettings,
RFModuleSettings,
SequencerSettings,
TimetagModuleSettings,
TimetagSequencerSettings,
)
from quantify_scheduler.enums import BinMode, TimeRef, TriggerCondition
from quantify_scheduler.instrument_coordinator.components import base
from quantify_scheduler.instrument_coordinator.utility import (
check_already_existing_acquisition,
lazy_set,
parameter_value_same_as_cache,
search_settable_param,
)
if TYPE_CHECKING:
from qblox_instruments.qcodes_drivers.module import Module
from qblox_instruments.qcodes_drivers.sequencer import Sequencer
from quantify_scheduler.schedules.schedule import (
AcquisitionMetadata,
CompiledSchedule,
)
[docs]
logger = logging.getLogger(__name__)
logger.setLevel(logging.WARNING)
# Prevent unsupported qblox-instruments version from crashing this submodule
driver_version_check.verify_qblox_instruments_version()
@dataclass(frozen=True)
[docs]
class _StaticHardwareProperties:
"""Dataclass for storing configuration differences across Qblox devices."""
[docs]
settings_type: type[BaseModuleSettings]
"""The settings dataclass to use that the hardware needs to configure to."""
[docs]
number_of_sequencers: int
"""The number of sequencers the hardware has available."""
[docs]
number_of_output_channels: int
"""The number of physical output channels that can be used."""
"""The number of physical input channels that can be used."""
@dataclass(frozen=True)
[docs]
class _StaticAnalogModuleProperties(_StaticHardwareProperties):
"""Dataclass for storing configuration differences across Qblox devices."""
[docs]
settings_type: type[AnalogModuleSettings]
"""The settings dataclass to use that the hardware needs to configure to."""
"""Specifies if an internal lo source is available."""
@dataclass(frozen=True)
[docs]
_QCM_BASEBAND_PROPERTIES = _StaticAnalogModuleProperties(
settings_type=AnalogModuleSettings,
has_internal_lo=False,
number_of_sequencers=constants.NUMBER_OF_SEQUENCERS_QCM,
number_of_output_channels=4,
number_of_input_channels=0,
)
[docs]
_QRM_BASEBAND_PROPERTIES = _StaticAnalogModuleProperties(
settings_type=AnalogModuleSettings,
has_internal_lo=False,
number_of_sequencers=constants.NUMBER_OF_SEQUENCERS_QRM,
number_of_output_channels=2,
number_of_input_channels=2,
)
[docs]
_QCM_RF_PROPERTIES = _StaticAnalogModuleProperties(
settings_type=RFModuleSettings,
has_internal_lo=True,
number_of_sequencers=constants.NUMBER_OF_SEQUENCERS_QCM,
number_of_output_channels=2,
number_of_input_channels=0,
)
[docs]
_QRM_RF_PROPERTIES = _StaticAnalogModuleProperties(
settings_type=RFModuleSettings,
has_internal_lo=True,
number_of_sequencers=constants.NUMBER_OF_SEQUENCERS_QRM,
number_of_output_channels=1,
number_of_input_channels=1,
)
[docs]
_QTM_PROPERTIES = _StaticTimetagModuleProperties(
settings_type=TimetagModuleSettings,
number_of_sequencers=constants.NUMBER_OF_SEQUENCERS_QTM,
number_of_output_channels=8,
number_of_input_channels=8,
)
[docs]
class _ModuleComponentBase(base.InstrumentCoordinatorComponentBase):
"""Qblox InstrumentCoordinator component base class."""
def __init__(self, instrument: Module) -> None:
super().__init__(instrument)
# The base class `InstrumentCoordinatorComponentBase` expects `instrument` to
# be a subclass of `Instrument`, _This_ class expects `instrument` to be a
# `Module` (for legacy reasons?), which does not subclass `Instrument` but
# `InstrumentChannel`, and is therefore not globally findable. Ergo, we store a
# reference here directly.
[docs]
self._instrument_module = instrument
[docs]
self._seq_name_to_idx_map = {
f"seq{idx}": idx for idx in range(self._hardware_properties.number_of_sequencers)
}
[docs]
self._nco_frequency_changed: dict[int, bool] = {}
"""
Private attribute for automatic mixer calibration. The keys are sequencer
indices. The `prepare` method resets this to an empty dictionary.
"""
# See the comment on self._instrument_module in __init__. Base class is incorrectly
# overridden, so we silence pyright.
@property
[docs]
def instrument(self) -> Module: # type: ignore
"""Returns a reference to the module instrument."""
return self._instrument_module
[docs]
def _set_parameter(
self,
instrument: Module | Sequencer,
parameter_name: str,
val: Any, # noqa: ANN401, disallow Any as type
) -> None:
"""
Set the parameter directly or using the lazy set.
Parameters
----------
instrument
The instrument or instrument channel that holds the parameter to set,
e.g. `self.instrument` or `self.instrument[f"sequencer{idx}"]`.
parameter_name
The name of the parameter to set.
val
The new value of the parameter.
"""
# TODO: these qcodes parameters already exist in the development branch
# of qblox-instruments, but will be released in 0.14.0 when RTP is
# officially supported (except bias tee).
# Until then, catching the value error is needed.
try:
search_settable_param(instrument=instrument, nested_parameter_name=parameter_name)
except ValueError as e:
if re.search(r".*(out|marker)[0-3]_bt_config", parameter_name) and val == "bypassed":
return
if re.search(
r".*(out|marker)[0-3]_bt_time_constant",
parameter_name,
):
return
raise e
if self.force_set_parameters():
instrument.set(parameter_name, val)
else:
lazy_set(instrument, parameter_name, val)
@property
[docs]
def is_running(self) -> bool:
"""
Finds if any of the sequencers is currently running.
Returns
-------
:
True if any of the sequencers reports the `SequencerStates.RUNNING` status.
"""
for seq_idx in range(self._hardware_properties.number_of_sequencers):
seq_status = self.instrument.get_sequencer_status(seq_idx)
if seq_status.state is SequencerStates.RUNNING:
return True
return False
[docs]
def wait_done(self, timeout_sec: int = 10) -> None:
"""
Blocks the instrument until all the sequencers are done running.
Parameters
----------
timeout_sec
The timeout in seconds. N.B. the instrument takes the timeout in minutes
(int), therefore it is rounded down to whole minutes with a minimum of 1.
"""
timeout_min = timeout_sec // 60
if timeout_min == 0:
timeout_min = 1
for idx in range(self._hardware_properties.number_of_sequencers):
state: SequencerStatus = self.instrument.get_sequencer_status(
sequencer=idx, timeout=timeout_min
)
for flag in state.info_flags:
logger.log(
level=logging.INFO,
msg=f"[{self.name}|seq{idx}] {flag} - {flag.value}",
)
for flag in state.warn_flags:
logger.log(
level=logging.WARNING,
msg=f"[{self.name}|seq{idx}] {flag} - {flag.value}",
)
for flag in state.err_flags:
logger.log(
level=logging.ERROR,
msg=f"[{self.name}|seq{idx}] {flag} - {flag.value}",
)
[docs]
def get_hardware_log(
self,
compiled_schedule: CompiledSchedule,
) -> dict | None:
"""
Retrieve the hardware log of the Qblox instrument associated to this component.
This log does not include the instrument serial number and firmware version.
Parameters
----------
compiled_schedule
Compiled schedule to check if this component is referenced in.
Returns
-------
:
A dict containing the hardware log of the Qblox instrument, in case the
component was referenced; else None.
"""
if self.instrument.name not in compiled_schedule.compiled_instructions:
return None
return _download_log(_get_configuration_manager(_get_instrument_ip(self)))
# Parameter name is different from base class. We ignore it because it is legacy
# code.
[docs]
def prepare(self, program: dict[str, dict]) -> None: # type: ignore
"""Store program containing sequencer settings."""
self._program = program
self._nco_frequency_changed = {}
[docs]
def disable_sync(self) -> None:
"""Disable sync for all sequencers."""
for idx in range(self._hardware_properties.number_of_sequencers):
# Prevent hanging on next run if instrument is not used.
self._set_parameter(self.instrument.sequencers[idx], "sync_en", False)
[docs]
def stop(self) -> None:
"""Stops all execution."""
self.disable_sync()
self.instrument.stop_sequencer()
@abstractmethod
[docs]
def arm_all_sequencers_in_program(self) -> None:
"""Arm all the sequencers that are part of the program."""
for seq_name in self._program.get("sequencers", {}):
if seq_name in self._seq_name_to_idx_map:
seq_idx = self._seq_name_to_idx_map[seq_name]
self.instrument.arm_sequencer(sequencer=seq_idx)
[docs]
def start(self) -> None:
"""Clear data, arm sequencers and start sequencers."""
self.clear_data()
self.arm_all_sequencers_in_program()
self._start_armed_sequencers()
[docs]
def _start_armed_sequencers(self) -> None:
"""Start execution of the schedule: start armed sequencers."""
for idx in range(self._hardware_properties.number_of_sequencers):
state = self.instrument.get_sequencer_status(idx)
if state.state is SequencerStates.ARMED:
self.instrument.start_sequencer(idx)
[docs]
def clear_data(self) -> None:
"""Clears remaining data on the module. Module type specific function."""
return None
@property
@abstractmethod
[docs]
def _hardware_properties(self) -> _StaticHardwareProperties:
"""
Holds all the differences between the different modules.
Returns
-------
:
A dataclass with all the hardware properties for this specific module.
"""
[docs]
class _AnalogModuleComponent(_ModuleComponentBase):
"""Qblox InstrumentCoordinator component base class."""
def __init__(self, instrument: Module) -> None:
super().__init__(instrument)
if instrument.is_rf_type is not self._hardware_properties.has_internal_lo:
raise RuntimeError(
f"{self.__class__.__name__} not compatible with the "
"provided instrument. Please confirm whether your device "
"is a Qblox RF or baseband module (having or not having an "
"internal LO)."
)
@abstractmethod
[docs]
def _determine_channel_map_parameters(
self, settings: AnalogSequencerSettings
) -> dict[str, str]:
"""Returns a dictionary with the channel map parameters for this module."""
channel_map_parameters = {}
self._determine_output_channel_map_parameters(settings, channel_map_parameters)
return channel_map_parameters
[docs]
def _determine_output_channel_map_parameters(
self, settings: AnalogSequencerSettings, channel_map_parameters: dict[str, str]
) -> dict[str, str]:
"""Adds the outputs to the channel map parameters dict."""
for channel_idx in range(self._hardware_properties.number_of_output_channels):
param_setting = "off"
# For baseband, output indices map 1-to-1 to channel map indices
if (
len(settings.connected_output_indices) > 0
and channel_idx in settings.connected_output_indices
and channel_idx in settings.connected_output_indices
):
if ChannelMode.COMPLEX in settings.channel_name:
param_setting = ["I", "Q", "I", "Q"][channel_idx]
elif ChannelMode.REAL in settings.channel_name:
param_setting = "I"
channel_map_parameters[f"connect_out{channel_idx}"] = param_setting
return channel_map_parameters
@property
@abstractmethod
[docs]
def _hardware_properties(self) -> _StaticAnalogModuleProperties:
"""
Holds all the differences between the different modules.
Returns
-------
:
A dataclass with all the hardware properties for this specific module.
"""
[docs]
class _QCMComponent(_AnalogModuleComponent):
"""QCM specific InstrumentCoordinator component."""
[docs]
_hardware_properties = _QCM_BASEBAND_PROPERTIES
def __init__(self, instrument: Module) -> None:
if not instrument.is_qcm_type:
raise TypeError(
f"Trying to create _QCMComponent from non-QCM instrument "
f'of type "{type(instrument)}".'
)
super().__init__(instrument)
[docs]
def retrieve_acquisition(self) -> None:
"""
Retrieves the previous acquisition.
Returns
-------
:
QCM returns None since the QCM has no acquisition.
"""
return None
[docs]
def prepare(self, program: dict[str, dict | AnalogModuleSettings]) -> None:
"""
Uploads the waveforms and programs to the sequencers.
All the settings that are required are configured. Keep in mind that
values set directly through the driver may be overridden (e.g. the
offsets will be set according to the specified mixer calibration
parameters).
Parameters
----------
program
Program to upload to the sequencers.
Under the key :code:`"sequencer"` you specify the sequencer specific
options for each sequencer, e.g. :code:`"seq0"`.
For global settings, the options are under different keys, e.g. :code:`"settings"`.
"""
super().prepare(program)
if (settings := program.get("settings")) is not None:
if isinstance(settings, dict):
settings = self._hardware_properties.settings_type.from_dict(settings)
self._configure_global_settings(settings)
for seq_idx in range(self._hardware_properties.number_of_sequencers):
self._set_parameter(self.instrument.sequencers[seq_idx], "sync_en", False)
for seq_name, settings in program["sequencers"].items():
if isinstance(settings, dict):
sequencer_settings = AnalogSequencerSettings.from_dict(settings)
else:
sequencer_settings = settings
if seq_name in self._seq_name_to_idx_map:
seq_idx = self._seq_name_to_idx_map[seq_name]
else:
raise KeyError(
f"Invalid program. Attempting to access non-existing sequencer "
f'with name "{seq_name}".'
)
self._configure_sequencer_settings(seq_idx=seq_idx, settings=sequencer_settings)
self._configure_nco_mixer_calibration(seq_idx=seq_idx, settings=sequencer_settings)
[docs]
class _QRMComponent(_AnalogModuleComponent):
"""QRM specific InstrumentCoordinator component."""
[docs]
_hardware_properties = _QRM_BASEBAND_PROPERTIES
def __init__(self, instrument: Module) -> None:
if not instrument.is_qrm_type:
raise TypeError(
f"Trying to create _QRMComponent from non-QRM instrument "
f'of type "{type(instrument)}".'
)
super().__init__(instrument)
[docs]
self._acquisition_manager: _QRMAcquisitionManager | None = None
"""Holds all the acquisition related logic."""
[docs]
def retrieve_acquisition(self) -> Dataset | None:
"""
Retrieves the latest acquisition results.
Returns
-------
:
The acquired data.
"""
if self._acquisition_manager:
return self._acquisition_manager.retrieve_acquisition()
else:
return None
[docs]
def prepare(self, program: dict[str, dict | BasebandModuleSettings]) -> None:
"""
Uploads the waveforms and programs to the sequencers.
All the settings that are required are configured. Keep in mind that
values set directly through the driver may be overridden (e.g. the
offsets will be set according to the specified mixer calibration
parameters).
Parameters
----------
program
Program to upload to the sequencers.
Under the key :code:`"sequencer"` you specify the sequencer specific
options for each sequencer, e.g. :code:`"seq0"`.
For global settings, the options are under different keys, e.g. :code:`"settings"`.
"""
super().prepare(program)
for seq_idx in range(self._hardware_properties.number_of_sequencers):
self._set_parameter(self.instrument.sequencers[seq_idx], "sync_en", False)
acq_duration = {}
for seq_name, settings in program["sequencers"].items():
if isinstance(settings, dict):
sequencer_settings = AnalogSequencerSettings.from_dict(settings)
else:
sequencer_settings = settings
if seq_name in self._seq_name_to_idx_map:
seq_idx = self._seq_name_to_idx_map[seq_name]
else:
raise KeyError(
f"Invalid program. Attempting to access non-existing sequencer "
f'with name "{seq_name}".'
)
self._configure_sequencer_settings(seq_idx=seq_idx, settings=sequencer_settings)
self._configure_nco_mixer_calibration(seq_idx=seq_idx, settings=sequencer_settings)
acq_duration[seq_name] = sequencer_settings.integration_length_acq
if (acq_metadata := program.get("acq_metadata")) is not None:
scope_mode_sequencer_and_qblox_acq_index = (
self._determine_scope_mode_acquisition_sequencer_and_qblox_acq_index(acq_metadata)
)
self._acquisition_manager = _QRMAcquisitionManager(
parent=self,
acquisition_metadata=acq_metadata,
scope_mode_sequencer_and_qblox_acq_index=scope_mode_sequencer_and_qblox_acq_index,
acquisition_duration=acq_duration,
seq_name_to_idx_map=self._seq_name_to_idx_map,
)
if scope_mode_sequencer_and_qblox_acq_index is not None:
self._set_parameter(
self.instrument,
"scope_acq_sequencer_select",
scope_mode_sequencer_and_qblox_acq_index[0],
)
else:
self._acquisition_manager = None
if (settings := program.get("settings")) is not None:
if isinstance(settings, dict):
settings = self._hardware_properties.settings_type.from_dict(settings)
self._configure_global_settings(settings)
for path in [
0,
1,
]:
self._set_parameter(self.instrument, f"scope_acq_trigger_mode_path{path}", "sequencer")
self._set_parameter(self.instrument, f"scope_acq_avg_mode_en_path{path}", True)
[docs]
def _determine_channel_map_parameters(
self, settings: AnalogSequencerSettings
) -> dict[str, str]:
"""Returns a dictionary with the channel map parameters for this module."""
channel_map_parameters = {}
self._determine_output_channel_map_parameters(settings, channel_map_parameters)
self._determine_input_channel_map_parameters(settings, channel_map_parameters)
return channel_map_parameters
[docs]
def _determine_scope_mode_acquisition_sequencer_and_qblox_acq_index(
self, acquisition_metadata: dict[str, AcquisitionMetadata]
) -> tuple[int, int] | None:
"""
Finds the sequencer and qblox_acq_index that performs the raw trace acquisition.
Raises an error if multiple scope mode acquisitions are present per sequencer.
Note, that compiler ensures there is at most one scope mode acquisition,
however the user is able to freely modify the compiler program,
so we make sure this requirement is still satisfied. See
:func:`~quantify_scheduler.backends.qblox.analog.AnalogModuleCompiler._ensure_single_scope_mode_acquisition_sequencer`.
Parameters
----------
acquisition_metadata
The acquisition metadata for each sequencer.
Returns
-------
:
The sequencer and qblox_acq_channel for the trace acquisition, if there is any,
otherwise None.
"""
sequencer_and_qblox_acq_index = None
for (
sequencer_name,
current_acquisition_metadata,
) in acquisition_metadata.items():
if current_acquisition_metadata.acq_protocol == "Trace":
# It's in the format "seq{n}", so we cut it.
sequencer_id = self._seq_name_to_idx_map[sequencer_name]
if (
sequencer_and_qblox_acq_index is not None
and sequencer_and_qblox_acq_index[0] != sequencer_id
):
single_scope_mode_acquisition_raise(
sequencer_0=sequencer_id,
sequencer_1=sequencer_and_qblox_acq_index[0],
module_name=self.name,
)
# For scope protocol, only one channel makes sense,
# we only need the first key in dict
qblox_acq_index = next(
iter(current_acquisition_metadata.acq_channels_metadata.keys())
)
sequencer_and_qblox_acq_index = (sequencer_id, qblox_acq_index)
return sequencer_and_qblox_acq_index
[docs]
def clear_data(self) -> None:
"""Clears remaining data on the module. Module type specific function."""
for sequencer_id in range(self._hardware_properties.number_of_sequencers):
self.instrument.delete_acquisition_data(sequencer=sequencer_id, all=True)
[docs]
class _RFComponent(_AnalogModuleComponent):
"""Mix-in for RF-module-specific InstrumentCoordinatorComponent behaviour."""
[docs]
def prepare(self, program: dict[str, dict | RFModuleSettings]) -> None:
"""
Uploads the waveforms and programs to the sequencers.
Overrides the parent method to additionally set LO settings for automatic mixer
calibration. This must be done _after_ all NCO frequencies have been set.
Parameters
----------
program
Program to upload to the sequencers.
Under the key :code:`"sequencer"` you specify the sequencer specific
options for each sequencer, e.g. :code:`"seq0"`.
For global settings, the options are under different keys, e.g. :code:`"settings"`.
"""
super().prepare(program)
lo_idx_to_connected_seq_idx: dict[int, list[int]] = {}
for seq_name, settings in program["sequencers"].items():
if isinstance(settings, dict):
sequencer_settings = AnalogSequencerSettings.from_dict(settings)
else:
sequencer_settings = settings
if seq_name in self._seq_name_to_idx_map:
seq_idx = self._seq_name_to_idx_map[seq_name]
else:
raise KeyError(
f"Invalid program. Attempting to access non-existing sequencer "
f'with name "{seq_name}".'
)
for lo_idx in self._get_connected_lo_idx_for_sequencer(
sequencer_settings=sequencer_settings
):
if lo_idx not in lo_idx_to_connected_seq_idx:
lo_idx_to_connected_seq_idx[lo_idx] = []
lo_idx_to_connected_seq_idx[lo_idx].append(seq_idx)
if (settings := program.get("settings")) is not None:
if isinstance(settings, dict):
settings = self._hardware_properties.settings_type.from_dict(settings)
self._configure_lo_settings(
settings=settings,
lo_idx_to_connected_seq_idx=lo_idx_to_connected_seq_idx,
)
[docs]
def _determine_output_channel_map_parameters(
self, settings: AnalogSequencerSettings, channel_map_parameters: dict[str, str]
) -> dict[str, str]:
"""Adds the outputs to the channel map parameters dict."""
expected_output_indices = {0: (0, 1), 1: (2, 3)}
for channel_idx in range(self._hardware_properties.number_of_output_channels):
param_setting = "off"
if (
ChannelMode.DIGITAL not in settings.channel_name
and len(settings.connected_output_indices) > 0
and tuple(settings.connected_output_indices)
== tuple(expected_output_indices[channel_idx])
):
param_setting = "IQ"
channel_map_parameters[f"connect_out{channel_idx}"] = param_setting
return channel_map_parameters
[docs]
def _get_connected_lo_idx_for_sequencer(
self, sequencer_settings: AnalogSequencerSettings
) -> list[int]:
"""
Looks at the connected _output_ ports of the sequencer (if any) to determine
which LO this sequencer's output is coupled to.
"""
connected_lo_idx = []
channel_map_parameters = self._determine_output_channel_map_parameters(
sequencer_settings, channel_map_parameters={}
)
for channel_idx in range(self._hardware_properties.number_of_output_channels):
if channel_map_parameters.get(f"connect_out{channel_idx}") == "IQ":
connected_lo_idx.append(channel_idx)
return connected_lo_idx
@abstractmethod
[docs]
class _QCMRFComponent(_RFComponent, _QCMComponent):
"""QCM-RF specific InstrumentCoordinator component."""
[docs]
_hardware_properties = _QCM_RF_PROPERTIES
[docs]
class _QRMRFComponent(_RFComponent, _QRMComponent):
"""QRM-RF specific InstrumentCoordinator component."""
[docs]
_hardware_properties = _QRM_RF_PROPERTIES
[docs]
class _QTMComponent(_ModuleComponentBase):
"""QTM specific InstrumentCoordinator component."""
[docs]
_hardware_properties = _QTM_PROPERTIES
def __init__(self, instrument: Module) -> None:
if not instrument.is_qtm_type:
raise TypeError(
f"Trying to create _QTMComponent from non-QTM instrument "
f'of type "{type(instrument)}".'
)
super().__init__(instrument)
[docs]
self._acquisition_manager: _QTMAcquisitionManager | None = None
"""Holds all the acquisition related logic."""
[docs]
def retrieve_acquisition(self) -> Dataset | None:
"""
Retrieves the latest acquisition results.
Returns
-------
:
The acquired data.
"""
if self._acquisition_manager:
return self._acquisition_manager.retrieve_acquisition()
else:
return None
[docs]
def prepare(self, program: dict[str, dict | TimetagModuleSettings]) -> None:
"""
Uploads the waveforms and programs to the sequencers.
All the settings that are required are configured. Keep in mind that
values set directly through the driver may be overridden (e.g. the
offsets will be set according to the specified mixer calibration
parameters).
Parameters
----------
program
Program to upload to the sequencers.
Under the key :code:`"sequencer"` you specify the sequencer specific
options for each sequencer, e.g. :code:`"seq0"`.
For global settings, the options are under different keys, e.g. :code:`"settings"`.
"""
super().prepare(program)
for seq_idx in range(self._hardware_properties.number_of_sequencers):
self._set_parameter(self.instrument.sequencers[seq_idx], "sync_en", False)
trace_acq_duration = {}
for seq_name, settings in program["sequencers"].items():
if seq_name in self._seq_name_to_idx_map:
seq_idx = self._seq_name_to_idx_map[seq_name]
else:
raise KeyError(
f"Invalid program. Attempting to access non-existing sequencer "
f'with name "{seq_name}".'
)
# 1-1 Sequencer-io_channel coupling,
# the io_channel settings are inside SequencerSettings
self._configure_sequencer_settings(seq_idx=seq_idx, settings=settings)
self._configure_io_channel_settings(seq_idx=seq_idx, settings=settings)
trace_acq_duration[seq_name] = settings.trace_acq_duration
if (acq_metadata := program.get("acq_metadata")) is not None:
self._acquisition_manager = _QTMAcquisitionManager(
parent=self,
acquisition_metadata=acq_metadata,
acquisition_duration=trace_acq_duration,
seq_name_to_idx_map=self._seq_name_to_idx_map,
)
else:
self._acquisition_manager = None
if (settings := program.get("settings")) is not None:
self._configure_global_settings(settings)
# No global settings yet.
[docs]
def clear_data(self) -> None:
"""Clears remaining data on the module. Module type specific function."""
for sequencer_id in range(self._hardware_properties.number_of_sequencers):
self.instrument.delete_acquisition_data(sequencer=sequencer_id, all=True)
[docs]
_ReadoutModuleComponentT = Union[_QRMComponent, _QTMComponent]
[docs]
class _AcquisitionManagerBase(ABC):
"""
Utility class that handles the acquisitions performed with a module.
An instance of this class is meant to exist only for a single prepare-start-
retrieve_acquisition cycle to prevent stateful behavior.
Parameters
----------
parent
Reference to the parent QRM IC component.
acquisition_metadata
Provides a summary of the used acquisition protocol, bin mode, acquisition channels,
acquisition indices per channel, and repetitions, for each sequencer.
acquisition_duration
The duration of each acquisition for each sequencer.
seq_name_to_idx_map
All available sequencer names to their ids in a dict.
"""
def __init__(
self,
parent: _ReadoutModuleComponentT,
acquisition_metadata: dict[str, AcquisitionMetadata],
acquisition_duration: dict[str, int],
seq_name_to_idx_map: dict[str, int],
) -> None:
[docs]
self._acq_duration = acquisition_duration
[docs]
self._seq_name_to_idx_map = seq_name_to_idx_map
@property
[docs]
def instrument(self) -> Module:
"""Returns the QRM driver from the parent IC component."""
return self.parent.instrument
@staticmethod
@abstractmethod
[docs]
def _check_bin_mode_compatible(acquisition_metadata: AcquisitionMetadata) -> None:
pass
@property
@abstractmethod
[docs]
def _protocol_to_acq_function_map(self) -> dict[str, Callable]:
"""
Mapping from acquisition protocol name to the function that processes the raw
acquisition data.
The acquisition processing function signature should be the following (for
brevity, it's not listed in the typehint):
.. code-block:: python
def acq_processing_function(
self,
acq_indices: list,
hardware_retrieved_acquisitions: dict,
acquisition_metadata: AcquisitionMetadata,
acq_duration: int,
qblox_acq_index: int,
acq_channel: Hashable,
) -> DataArray:
"""
[docs]
def retrieve_acquisition(self) -> Dataset:
"""
Retrieves all the acquisition data in the correct format.
Returns
-------
:
The acquisitions with the protocols specified in the `acquisition_metadata`.
Each `xarray.DataArray` in the `xarray.Dataset` corresponds to one `acq_channel`.
The ``acq_channel`` is the name of each `xarray.DataArray` in the `xarray.Dataset`.
Each `xarray.DataArray` is a two-dimensional array, with ``acq_index`` and
Each `xarray.DataArray` is a two-dimensional array,
with ``acq_index`` and ``repetition`` as dimensions.
"""
dataset = Dataset()
for sequencer_name, acquisition_metadata in self._acquisition_metadata.items():
self._check_bin_mode_compatible(acquisition_metadata)
acquisition_function: Callable = self._protocol_to_acq_function_map[
acquisition_metadata.acq_protocol
]
# retrieve the raw data from the qrm sequencer module
hardware_retrieved_acquisitions = self._get_acquisitions_from_instrument(
seq_idx=self._seq_name_to_idx_map[sequencer_name],
acquisition_metadata=acquisition_metadata,
)
for (
qblox_acq_index,
acq_channel_metadata,
) in acquisition_metadata.acq_channels_metadata.items():
acq_channel: Hashable = acq_channel_metadata.acq_channel
acq_indices: list[int] = acq_channel_metadata.acq_indices
self._assert_acquisition_data_exists(
hardware_retrieved_acquisitions, qblox_acq_index, acq_channel
)
# the acquisition_function retrieves the right part of the acquisitions
# data structure returned by the qrm
formatted_acquisitions = acquisition_function(
acq_indices=acq_indices,
hardware_retrieved_acquisitions=hardware_retrieved_acquisitions,
acquisition_metadata=acquisition_metadata,
acq_duration=self._acq_duration[sequencer_name],
qblox_acq_index=qblox_acq_index,
acq_channel=acq_channel,
)
formatted_acquisitions_dataset = Dataset({acq_channel: formatted_acquisitions})
check_already_existing_acquisition(
new_dataset=formatted_acquisitions_dataset, current_dataset=dataset
)
dataset = dataset.merge(formatted_acquisitions_dataset)
return dataset
[docs]
def _assert_acquisition_data_exists(
self,
hardware_retrieved_acquisitions: dict,
qblox_acq_index: int,
acq_channel: Hashable,
) -> None:
"""Assert that the qblox_acq_index is in the acquisition data."""
qblox_acq_name = self._qblox_acq_index_to_qblox_acq_name(qblox_acq_index)
if qblox_acq_name not in hardware_retrieved_acquisitions:
raise KeyError(
f"The acquisition data retrieved from the hardware does not contain "
f"data for acquisition channel {acq_channel} (referred to by Qblox "
f"acquisition index {qblox_acq_index}).\n"
f"{hardware_retrieved_acquisitions=}"
)
[docs]
def _get_acquisitions_from_instrument(
self,
seq_idx: int,
acquisition_metadata: AcquisitionMetadata, # noqa: ARG002, unused parameter
) -> dict:
return self.instrument.get_acquisitions(seq_idx)
@staticmethod
[docs]
def _acq_channel_attrs(
protocol: str,
) -> dict:
return {"acq_protocol": protocol}
@classmethod
[docs]
def _get_bin_data(cls, hardware_retrieved_acquisitions: dict, qblox_acq_index: int = 0) -> dict:
"""Returns the bin entry of the acquisition data dict."""
qblox_acq_name = cls._qblox_acq_index_to_qblox_acq_name(qblox_acq_index)
channel_data = hardware_retrieved_acquisitions[qblox_acq_name]
if channel_data["index"] != qblox_acq_index:
raise RuntimeError(
f"Name does not correspond to a valid acquisition for name {qblox_acq_name}, "
f'which has index {channel_data["index"]}.'
)
return channel_data["acquisition"]["bins"]
@staticmethod
[docs]
def _qblox_acq_index_to_qblox_acq_name(qblox_acq_index: int) -> str:
"""Returns the name of the acquisition from the qblox_acq_index."""
return str(qblox_acq_index)
[docs]
def _get_trigger_count_threshold_data(
self,
*,
acq_indices: list, # noqa: ARG002, unused argument
hardware_retrieved_acquisitions: dict,
acquisition_metadata: AcquisitionMetadata,
acq_duration: int, # noqa: ARG002, unused argument
qblox_acq_index: int,
acq_channel: Hashable,
) -> DataArray:
"""
Retrieve the thresholded acquisition data associated with ``acq_channel`` and ``acq_index``.
Parameters
----------
acq_indices
Acquisition indices.
hardware_retrieved_acquisitions
The acquisitions dict as returned by the sequencer.
acquisition_metadata
Acquisition metadata.
acq_duration
Desired maximum number of samples for the scope acquisition.
qblox_acq_index
The Qblox acquisition index from which to get the data.
acq_channel
The acquisition channel.
Returns
-------
:
DataArray containing thresholded acquisition data.
"""
bin_data = self._get_bin_data(hardware_retrieved_acquisitions, qblox_acq_index)
acq_index_dim_name = f"acq_index_{acq_channel}"
thresh_tc_settings = acquisition_metadata.acq_channels_metadata[
qblox_acq_index
].thresholded_trigger_count
assert thresh_tc_settings is not None
# "avg_cnt" is the key for QRM modules, "count" for QTM modules. Note
# that QTM modules also return a "avg_cnt" key, that should not be used!
counts = bin_data["count"] if "count" in bin_data else bin_data["avg_cnt"]
# To make the return data similar to thresholdd acquisition, it is cast
# to integers. Note that the hardwware returns counts as floats.
if thresh_tc_settings.condition == TriggerCondition.GREATER_THAN_EQUAL_TO:
states = (np.round(counts) >= thresh_tc_settings.threshold).astype(int)
elif thresh_tc_settings.condition == TriggerCondition.LESS_THAN:
states = (np.round(counts) < thresh_tc_settings.threshold).astype(int)
else:
raise ValueError(f"Unknown trigger condition {thresh_tc_settings.condition}")
# The above lines convert NaNs to 0. We want them to be -1.
states[np.isnan(counts)] = -1
return DataArray(
[states],
dims=["repetition", acq_index_dim_name],
coords={"repetition": [0], acq_index_dim_name: range(len(states))},
attrs=self._acq_channel_attrs(acquisition_metadata.acq_protocol),
)
[docs]
class _QRMAcquisitionManager(_AcquisitionManagerBase):
"""
Utility class that handles the acquisitions performed with the QRM.
An instance of this class is meant to exist only for a single prepare-start-
retrieve_acquisition cycle to prevent stateful behavior.
Parameters
----------
parent
Reference to the parent QRM IC component.
acquisition_metadata
Provides a summary of the used acquisition protocol, bin mode, acquisition channels,
acquisition indices per channel, and repetitions, for each sequencer.
acquisition_duration
The duration of each acquisition for each sequencer.
seq_name_to_idx_map
All available sequencer names to their ids in a dict.
scope_mode_sequencer_and_qblox_acq_index
The sequencer and qblox acq_index of the scope mode acquisition if there's any.
"""
def __init__(
self,
parent: _QRMComponent,
acquisition_metadata: dict[str, AcquisitionMetadata],
acquisition_duration: dict[str, int],
seq_name_to_idx_map: dict[str, int],
scope_mode_sequencer_and_qblox_acq_index: tuple[int, int] | None = None,
) -> None:
super().__init__(
parent=parent,
acquisition_metadata=acquisition_metadata,
acquisition_duration=acquisition_duration,
seq_name_to_idx_map=seq_name_to_idx_map,
)
[docs]
self._scope_mode_sequencer_and_qblox_acq_index = scope_mode_sequencer_and_qblox_acq_index
@property
[docs]
def _protocol_to_acq_function_map(self) -> dict[str, Callable]:
return {
"WeightedIntegratedSeparated": partial(self._get_integration_data, separated=True),
"NumericalSeparatedWeightedIntegration": partial(
self._get_integration_data, separated=True
),
"NumericalWeightedIntegration": partial(self._get_integration_data, separated=False),
"SSBIntegrationComplex": self._get_integration_amplitude_data,
"ThresholdedAcquisition": self._get_threshold_data,
"ThresholdedTriggerCount": self._get_trigger_count_threshold_data,
"Trace": self._get_scope_data,
"TriggerCount": self._get_trigger_count_data,
}
@staticmethod
[docs]
def _check_bin_mode_compatible(acquisition_metadata: AcquisitionMetadata) -> None:
if (
acquisition_metadata.bin_mode
not in QRM_COMPATIBLE_BIN_MODES[acquisition_metadata.acq_protocol]
):
raise IncompatibleBinModeError(
module_type="QRM",
protocol=acquisition_metadata.acq_protocol,
bin_mode=acquisition_metadata.bin_mode,
)
[docs]
def retrieve_acquisition(self) -> Dataset:
"""
Retrieves all the acquisition data in the correct format.
Returns
-------
:
The acquisitions with the protocols specified in the `acquisition_metadata`.
Each `xarray.DataArray` in the `xarray.Dataset` corresponds to one `acq_channel`.
The ``acq_channel`` is the name of each `xarray.DataArray` in the `xarray.Dataset`.
Each `xarray.DataArray` is a two-dimensional array,
with ``acq_index`` and ``repetition`` as dimensions.
"""
self._store_scope_acquisition()
return super().retrieve_acquisition()
[docs]
def _store_scope_acquisition(self) -> None:
"""
Calls :code:`store_scope_acquisition` function on the Qblox instrument.
This will ensure that the correct sequencer will store the scope acquisition
data on the hardware, so it will be filled out when we call :code:`get_acquisitions`
on the Qblox instrument's sequencer corresponding to the scope acquisition.
"""
if self._scope_mode_sequencer_and_qblox_acq_index is None:
return
sequencer_index = self._scope_mode_sequencer_and_qblox_acq_index[0]
if sequencer_index not in self._seq_name_to_idx_map.values():
raise ValueError(
f"Attempting to retrieve scope mode data from sequencer "
f"{sequencer_index}. A QRM only has the following sequencer indices: "
f"{list(self._seq_name_to_idx_map.values())}."
)
qblox_acq_index = self._scope_mode_sequencer_and_qblox_acq_index[1]
qblox_acq_name = self._qblox_acq_index_to_qblox_acq_name(qblox_acq_index)
self.instrument.store_scope_acquisition(sequencer_index, qblox_acq_name)
[docs]
def _get_scope_data(
self,
*,
acq_indices: list,
hardware_retrieved_acquisitions: dict,
acquisition_metadata: AcquisitionMetadata,
acq_duration: int,
qblox_acq_index: int,
acq_channel: Hashable,
) -> DataArray:
"""
Retrieves the scope mode acquisition associated with an `acq_channel`.
Parameters
----------
acq_indices
Acquisition indices.
hardware_retrieved_acquisitions
The acquisitions dict as returned by the sequencer.
acquisition_metadata
Acquisition metadata.
acq_duration
Desired maximum number of samples for the scope acquisition.
qblox_acq_index
The Qblox acquisition index from which to get the data.
acq_channel
The acquisition channel.
Returns
-------
:
The scope mode data.
"""
if acq_duration < 0 or acq_duration > constants.MAX_SAMPLE_SIZE_SCOPE_ACQUISITIONS:
raise ValueError(
"Attempting to retrieve sample of size "
f"{acq_duration}, but only integer values "
f"0,...,{constants.MAX_SAMPLE_SIZE_SCOPE_ACQUISITIONS} "
f"are allowed."
)
qblox_acq_name = self._qblox_acq_index_to_qblox_acq_name(qblox_acq_index)
scope_data = hardware_retrieved_acquisitions[qblox_acq_name]["acquisition"]["scope"]
for path_label in ("path0", "path1"):
if scope_data[path_label]["out-of-range"]:
logger.warning(
f"The scope mode data of {path_label} of {self.parent.name} with "
f"acq_channel={acq_channel} was out-of-range."
)
# NB hardware already divides by avg_count for scope mode
scope_data_i = np.array(scope_data["path0"]["data"][:acq_duration])
scope_data_q = np.array(scope_data["path1"]["data"][:acq_duration])
acq_index_dim_name = f"acq_index_{acq_channel}"
trace_index_dim_name = f"trace_index_{acq_channel}"
return DataArray(
(scope_data_i + scope_data_q * 1j).reshape((1, -1)),
dims=[acq_index_dim_name, trace_index_dim_name],
coords={
acq_index_dim_name: acq_indices,
trace_index_dim_name: list(range(acq_duration)),
},
attrs=self._acq_channel_attrs(acquisition_metadata.acq_protocol),
)
[docs]
def _get_integration_data(
self,
*,
acq_indices: list,
hardware_retrieved_acquisitions: dict,
acquisition_metadata: AcquisitionMetadata,
acq_duration: int, # noqa: ARG002, unused argument
qblox_acq_index: int,
acq_channel: Hashable,
multiplier: float = 1,
separated: bool = True,
) -> DataArray:
"""
Retrieves the integrated acquisition data associated with an `acq_channel`.
Parameters
----------
acq_indices
Acquisition indices.
hardware_retrieved_acquisitions
The acquisitions dict as returned by the sequencer.
acquisition_metadata
Acquisition metadata.
acq_duration
Desired maximum number of samples for the scope acquisition.
qblox_acq_index
The Qblox acquisition index from which to get the data.
acq_channel
The acquisition channel.
multiplier
Multiplies the data with this number.
separated
True: return I and Q data separately
False: return I+Q in the real part and 0 in the imaginary part
Returns
-------
:
The integrated data.
"""
bin_data = self._get_bin_data(hardware_retrieved_acquisitions, qblox_acq_index)
i_data = np.array(bin_data["integration"]["path0"])
q_data = np.array(bin_data["integration"]["path1"])
if not separated:
i_data = i_data + q_data
q_data = np.zeros_like(q_data)
acquisitions_data = multiplier * (i_data + q_data * 1j)
acq_index_dim_name = f"acq_index_{acq_channel}"
if acquisition_metadata.bin_mode == BinMode.AVERAGE:
return DataArray(
acquisitions_data.reshape((len(acq_indices),)),
dims=[acq_index_dim_name],
coords={acq_index_dim_name: acq_indices},
attrs=self._acq_channel_attrs(acquisition_metadata.acq_protocol),
)
elif acquisition_metadata.repetitions * len(acq_indices) == acquisitions_data.size:
acq_data = acquisitions_data.reshape(
(acquisition_metadata.repetitions, len(acq_indices))
)
return DataArray(
acq_data,
dims=["repetition", acq_index_dim_name],
coords={acq_index_dim_name: acq_indices},
attrs=self._acq_channel_attrs(acquisition_metadata.acq_protocol),
)
# There is control flow containing measurements, skip reshaping
else:
warnings.warn(
"The format of acquisition data of looped measurements in APPEND mode"
" will change in a future quantify-scheduler revision.",
FutureWarning,
)
acq_data = acquisitions_data.reshape((acquisition_metadata.repetitions, -1))
return DataArray(
acq_data,
dims=["repetition", "loop_repetition"],
coords=None,
attrs=self._acq_channel_attrs(acquisition_metadata.acq_protocol),
)
[docs]
def _get_integration_amplitude_data(
self,
acq_indices: list,
hardware_retrieved_acquisitions: dict,
acquisition_metadata: AcquisitionMetadata,
acq_duration: int,
qblox_acq_index: int,
acq_channel: Hashable,
) -> DataArray:
"""
Gets the integration data but normalized to the integration time.
The return value is thus the amplitude of the demodulated
signal directly and has volt units (i.e. same units as a single sample of the
integrated signal).
Parameters
----------
acq_indices
Acquisition indices.
hardware_retrieved_acquisitions
The acquisitions dict as returned by the sequencer.
acquisition_metadata
Acquisition metadata.
acq_duration
Desired maximum number of samples for the scope acquisition.
qblox_acq_index
The Qblox acquisition index from which to get the data.
acq_channel
The acquisition channel.
Returns
-------
:
Array containing binned, normalized acquisition data.
"""
if acq_duration is None:
raise RuntimeError(
"Retrieving data failed. Expected the integration length to be defined,"
" but it is `None`."
)
formatted_data = self._get_integration_data(
acq_indices=acq_indices,
hardware_retrieved_acquisitions=hardware_retrieved_acquisitions,
acquisition_metadata=acquisition_metadata,
acq_duration=acq_duration,
qblox_acq_index=qblox_acq_index,
acq_channel=acq_channel,
multiplier=1 / acq_duration,
)
return formatted_data
[docs]
def _get_threshold_data(
self,
*,
acq_indices: list,
hardware_retrieved_acquisitions: dict,
acquisition_metadata: AcquisitionMetadata,
acq_duration: int,
qblox_acq_index: int,
acq_channel: Hashable,
) -> DataArray:
"""
Retrieve the thresholded acquisition data associated with ``acq_channel`` and ``acq_index``.
Parameters
----------
acq_indices
Acquisition indices.
hardware_retrieved_acquisitions
The acquisitions dict as returned by the sequencer.
acquisition_metadata
Acquisition metadata.
acq_duration
Desired maximum number of samples for the scope acquisition.
qblox_acq_index
The Qblox acquisition index from which to get the data.
acq_channel
The acquisition channel.
Returns
-------
:
DataArray containing thresholded acquisition data.
"""
if acq_duration is None:
raise RuntimeError(
"Retrieving data failed. Expected the integration length to be defined,"
" but it is `None`."
)
bin_data = self._get_bin_data(
hardware_retrieved_acquisitions=hardware_retrieved_acquisitions,
qblox_acq_index=qblox_acq_index,
)
acq_index_dim_name = f"acq_index_{acq_channel}"
acquisitions_data = np.array(
list(map(lambda n: n if not isnan(n) else -1, bin_data["threshold"])),
dtype=acquisition_metadata.acq_return_type,
)
if acquisition_metadata.bin_mode == BinMode.AVERAGE:
return DataArray(
acquisitions_data.reshape((len(acq_indices),)),
dims=[acq_index_dim_name],
coords={acq_index_dim_name: acq_indices},
attrs=self._acq_channel_attrs(acquisition_metadata.acq_protocol),
)
else:
return DataArray(
acquisitions_data.reshape((acquisition_metadata.repetitions, len(acq_indices))),
dims=["repetition", acq_index_dim_name],
coords={acq_index_dim_name: acq_indices},
attrs=self._acq_channel_attrs(acquisition_metadata.acq_protocol),
)
[docs]
def _get_trigger_count_data(
self,
*,
acq_indices: list, # noqa: ARG002 unused argument
hardware_retrieved_acquisitions: dict,
acquisition_metadata: AcquisitionMetadata,
acq_duration: int, # noqa: ARG002 unused argument
qblox_acq_index: int,
acq_channel: Hashable,
) -> DataArray:
"""
Retrieves the trigger count acquisition data associated with `acq_channel`.
Parameters
----------
acq_indices
Acquisition indices.
hardware_retrieved_acquisitions
The acquisitions dict as returned by the sequencer.
acquisition_metadata
Acquisition metadata.
acq_duration
Desired maximum number of samples for the scope acquisition.
qblox_acq_index
The Qblox acquisition index from which to get the data.
acq_channel
The acquisition channel.
Returns
-------
data : xarray.DataArray
The acquired trigger count data.
Notes
-----
- For BinMode.DISTRIBUTION, `data` contains the distribution of counts.
- For BinMode.APPEND, `data` contains the raw trigger counts.
"""
bin_data = self._get_bin_data(hardware_retrieved_acquisitions, qblox_acq_index)
acq_index_dim_name = f"acq_index_{acq_channel}"
if acquisition_metadata.bin_mode == BinMode.DISTRIBUTION:
def _convert_from_cumulative(
cumulative_values: list[int],
) -> dict[int, int]:
"""
Return the distribution of counts from a cumulative distribution.
Note, the cumulative distribution is in reverse order.
The cumulative_values list can contain any number of integers and NaNs.
"""
cumulative_values = list(enumerate(cumulative_values))
result = {}
last_cumulative_value = 0
for count, current_cumulative_value in reversed(cumulative_values): # type: ignore
if (not isnan(current_cumulative_value)) and (
last_cumulative_value != current_cumulative_value
):
result[count + 1] = current_cumulative_value - last_cumulative_value
last_cumulative_value = current_cumulative_value
return result
result = _convert_from_cumulative(bin_data["avg_cnt"])
return DataArray(
[list(result.values())[::-1]],
dims=["repetition", "counts"],
coords={"repetition": [0], "counts": list(result.keys())[::-1]},
attrs=self._acq_channel_attrs(acquisition_metadata.acq_protocol),
)
elif acquisition_metadata.bin_mode == BinMode.SUM:
counts = np.array(bin_data["avg_cnt"]).astype(int)
return DataArray(
counts.reshape((len(acq_indices),)),
dims=[acq_index_dim_name],
coords={acq_index_dim_name: acq_indices},
attrs=self._acq_channel_attrs(acquisition_metadata.acq_protocol),
)
elif acquisition_metadata.bin_mode == BinMode.APPEND:
counts = np.array(bin_data["avg_cnt"]).astype(int)
return DataArray(
[counts],
dims=["repetition", acq_index_dim_name],
coords={"repetition": [0], acq_index_dim_name: range(len(counts))},
attrs=self._acq_channel_attrs(acquisition_metadata.acq_protocol),
)
else:
# In principle unreachable due to _check_bin_mode_compatible, but included for
# completeness.
assert False, "This should not be reachable due to _check_bin_mode_compatible."
[docs]
class _QTMAcquisitionManager(_AcquisitionManagerBase):
"""
Utility class that handles the acquisitions performed with the QTM.
An instance of this class is meant to exist only for a single prepare-start-
retrieve_acquisition cycle to prevent stateful behavior.
Parameters
----------
parent
Reference to the parent QRM IC component.
acquisition_metadata
Provides a summary of the used acquisition protocol, bin mode, acquisition channels,
acquisition indices per channel, and repetitions, for each sequencer.
acquisition_duration
The duration of each acquisition for each sequencer.
seq_name_to_idx_map
All available sequencer names to their ids in a dict.
"""
@property
[docs]
def _protocol_to_acq_function_map(self) -> dict[str, Callable]:
return {
"TriggerCount": self._get_trigger_count_data,
"ThresholdedTriggerCount": self._get_trigger_count_threshold_data,
"Timetag": self._get_timetag_data,
"Trace": self._get_digital_trace_data,
"TimetagTrace": self._get_timetag_trace_data,
}
@staticmethod
[docs]
def _check_bin_mode_compatible(acquisition_metadata: AcquisitionMetadata) -> None:
if (
acquisition_metadata.bin_mode
not in QTM_COMPATIBLE_BIN_MODES[acquisition_metadata.acq_protocol]
):
raise IncompatibleBinModeError(
module_type="QTM",
protocol=acquisition_metadata.acq_protocol,
bin_mode=acquisition_metadata.bin_mode,
)
[docs]
def _get_acquisitions_from_instrument(
self, seq_idx: int, acquisition_metadata: AcquisitionMetadata
) -> dict:
data = super()._get_acquisitions_from_instrument(seq_idx, acquisition_metadata)
if acquisition_metadata.acq_protocol in ("Trace", "TimetagTrace"):
# We add this scope data in the same format as QRM acquisitions.
scope_data = self.instrument[f"io_channel{seq_idx}"].get_scope_data()
# For (timetag)trace acquisitions, there is only one acq channel per
# sequencer/io_channel (enforced by compiler). We just take the first one.
qblox_acq_index = next(iter(acquisition_metadata.acq_channels_metadata.keys()))
qblox_acq_name = self._qblox_acq_index_to_qblox_acq_name(qblox_acq_index)
data[qblox_acq_name]["acquisition"]["scope"] = scope_data
return data
[docs]
def _get_trigger_count_data(
self,
*,
acq_indices: list, # noqa: ARG002, unused argument
hardware_retrieved_acquisitions: dict,
acquisition_metadata: AcquisitionMetadata,
acq_duration: int, # noqa: ARG002, unused argument
qblox_acq_index: int,
acq_channel: Hashable,
) -> DataArray:
"""
Retrieves the trigger count acquisition data associated with `acq_channel`.
Parameters
----------
acq_indices
Acquisition indices.
hardware_retrieved_acquisitions
The acquisitions dict as returned by the sequencer.
acquisition_metadata
Acquisition metadata.
acq_duration
Desired maximum number of samples for the scope acquisition.
qblox_acq_index
The Qblox acquisition index from which to get the data.
acq_channel
The acquisition channel.
Returns
-------
data : xarray.DataArray
The acquired trigger count data.
Notes
-----
- BinMode.AVERAGE is not implemented for the QTM.
- For BinMode.APPEND, `data` contains the raw trigger counts.
"""
bin_data = self._get_bin_data(hardware_retrieved_acquisitions, qblox_acq_index)
acq_index_dim_name = f"acq_index_{acq_channel}"
counts = np.array(bin_data["count"]).astype(int)
if acquisition_metadata.bin_mode == BinMode.APPEND:
return DataArray(
[counts],
dims=["repetition", acq_index_dim_name],
coords={"repetition": [0], acq_index_dim_name: range(len(counts))},
attrs=self._acq_channel_attrs(acquisition_metadata.acq_protocol),
)
elif acquisition_metadata.bin_mode == BinMode.SUM:
return DataArray(
counts.reshape((len(acq_indices),)),
dims=[acq_index_dim_name],
coords={acq_index_dim_name: acq_indices},
attrs=self._acq_channel_attrs(acquisition_metadata.acq_protocol),
)
else:
# In principle unreachable due to _check_bin_mode_compatible, but included for
# completeness.
assert False, "This should not be reachable due to _check_bin_mode_compatible."
[docs]
def _get_digital_trace_data(
self,
*,
acq_indices: list,
hardware_retrieved_acquisitions: dict,
acquisition_metadata: AcquisitionMetadata,
acq_duration: int,
qblox_acq_index: int,
acq_channel: Hashable,
) -> DataArray:
qblox_acq_name = self._qblox_acq_index_to_qblox_acq_name(qblox_acq_index)
scope_data = np.array(
hardware_retrieved_acquisitions[qblox_acq_name]["acquisition"]["scope"][:acq_duration]
)
acq_index_dim_name = f"acq_index_{acq_channel}"
trace_index_dim_name = f"trace_index_{acq_channel}"
return DataArray(
scope_data.reshape((1, -1)),
dims=[acq_index_dim_name, trace_index_dim_name],
coords={
acq_index_dim_name: acq_indices,
trace_index_dim_name: list(range(acq_duration)),
},
attrs=self._acq_channel_attrs(acquisition_metadata.acq_protocol),
)
[docs]
_ClusterModule = Union[
_QCMComponent, _QRMComponent, _QCMRFComponent, _QRMRFComponent, _QTMComponent
]
"""Type that combines all the possible modules for a cluster."""
[docs]
class ClusterComponent(base.InstrumentCoordinatorComponentBase):
"""
Class that represents an instrument coordinator component for a Qblox cluster.
New instances of the ClusterComponent will automatically add installed
modules using name `"<cluster_name>_module<slot>"`.
Parameters
----------
instrument
Reference to the cluster driver object.
"""
def __init__(self, instrument: Cluster) -> None:
super().__init__(instrument)
[docs]
self._cluster_modules: dict[str, _ClusterModule] = {}
# Important: a tuple with only False may not occur as a key, because new
# unsupported module types may return False on all is_..._type functions.
module_type_map = {
(True, False, False, False): _QCMComponent,
(True, False, True, False): _QCMRFComponent,
(False, True, False, False): _QRMComponent,
(False, True, True, False): _QRMRFComponent,
(False, False, False, True): _QTMComponent,
}
for instrument_module in instrument.modules:
try:
icc_class: type = module_type_map[
(
instrument_module.is_qcm_type,
instrument_module.is_qrm_type,
instrument_module.is_rf_type,
getattr(instrument_module, "is_qtm_type", False),
)
]
except KeyError:
continue
self._cluster_modules[instrument_module.name] = icc_class(instrument_module)
@property
[docs]
def is_running(self) -> bool:
"""Returns true if any of the modules are currently running."""
return any(comp.is_running for comp in self._cluster_modules.values())
[docs]
def start(self) -> None:
"""Starts all the modules in the cluster."""
# Disarming all sequencers, to make sure the last
# `self.instrument.start_sequencer` only starts sequencers
# which are explicitly armed by the subsequent calls.
self.instrument.stop_sequencer()
# Arming all sequencers in the program.
for comp_name, comp in self._cluster_modules.items():
if comp_name in self._program:
comp.clear_data()
comp.arm_all_sequencers_in_program()
# Starts all sequencers in the cluster, time efficiently.
self.instrument.start_sequencer()
[docs]
def stop(self) -> None:
"""Stops all the modules in the cluster."""
for comp in self._cluster_modules.values():
comp.disable_sync()
# Stops all sequencers in the cluster, time efficiently.
self.instrument.stop_sequencer()
[docs]
def prepare(self, options: dict[str, dict]) -> None:
"""
Prepares the cluster component for execution of a schedule.
Parameters
----------
options
The compiled instructions to configure the cluster to.
"""
self._program = copy.deepcopy(options)
for name, comp_options in self._program.items():
if name == "settings":
self._configure_cmm_settings(settings=comp_options)
elif name in self._cluster_modules:
self._cluster_modules[name].prepare(comp_options)
else:
raise KeyError(
f"Attempting to prepare module {name} of cluster {self.name}, while"
f" module has not been added to the cluster component."
)
[docs]
def retrieve_acquisition(self) -> Dataset | None:
"""
Retrieves all the data from the instruments.
Returns
-------
:
The acquired data or ``None`` if no acquisitions have been performed.
"""
acquisitions = Dataset()
for comp_name, comp in self._cluster_modules.items():
if comp_name not in self._program:
continue
comp_acq = comp.retrieve_acquisition()
if comp_acq is not None:
check_already_existing_acquisition(
new_dataset=comp_acq, current_dataset=acquisitions
)
acquisitions.update(comp_acq)
return acquisitions if len(acquisitions) > 0 else None
[docs]
def wait_done(self, timeout_sec: int = 10) -> None:
"""
Blocks until all the components are done executing their programs.
Parameters
----------
timeout_sec
The time in seconds until the instrument is considered to have timed out.
"""
for comp in self._cluster_modules.values():
comp.wait_done(timeout_sec=timeout_sec)
[docs]
def get_hardware_log(
self,
compiled_schedule: CompiledSchedule,
) -> dict | None:
"""
Retrieve the hardware log of the Cluster Management Module and associated modules.
This log includes the module serial numbers and
firmware version.
Parameters
----------
compiled_schedule
Compiled schedule to check if this cluster is referenced in (and if so,
which specific modules are referenced in).
Returns
-------
:
A dict containing the hardware log of the cluster, in case the
component was referenced; else None.
"""
cluster = self.instrument
if cluster.name not in compiled_schedule.compiled_instructions:
return None
cluster_ip = _get_instrument_ip(self)
hardware_log = {
f"{cluster.name}_cmm": _download_log(
config_manager=_get_configuration_manager(cluster_ip),
is_cluster=True,
),
f"{cluster.name}_idn": str(cluster.get_idn()),
f"{cluster.name}_mods_info": str(cluster._get_mods_info()),
}
for module in cluster.modules:
if module.name in compiled_schedule.compiled_instructions[cluster.name]:
# Cannot fetch log from module.get_hardware_log here since modules are
# not InstrumentCoordinator components when using a cluster
module_ip = f"{cluster_ip}/{module.slot_idx}"
hardware_log[module.name] = _download_log(_get_configuration_manager(module_ip))
return hardware_log
[docs]
def _get_instrument_ip(component: base.InstrumentCoordinatorComponentBase) -> str:
ip_config = component.instrument.get_ip_config()
if ip_config == "0":
raise ValueError(
f"Instrument '{component.instrument.name}' returned {ip_config=}."
f"Please make sure the physical instrument is connected and has a valid ip."
)
instrument_ip = ip_config
if "/" in instrument_ip:
instrument_ip = instrument_ip.split("/")[0]
return instrument_ip
[docs]
def _get_configuration_manager(instrument_ip: str) -> ConfigurationManager:
try:
config_manager = ConfigurationManager(instrument_ip)
except RuntimeError as error:
new_message = f"{error}\nNote: qblox-instruments might have changed ip formatting."
raise type(error)(new_message)
return config_manager
[docs]
def _download_log(
config_manager: ConfigurationManager,
is_cluster: bool | None = False,
) -> dict:
hardware_log = {}
sources = ["app", "system"]
if is_cluster:
sources.append("cfg_man")
for source in sources:
# uuid prevents unwanted deletion if file already exists
temp_log_file_name = os.path.join(get_datadir(), f"{source}_{uuid4()}")
config_manager.download_log(source=source, fmt="txt", file=temp_log_file_name)
if os.path.isfile(temp_log_file_name):
with open(temp_log_file_name, encoding="utf-8", errors="replace") as file:
log = file.read()
os.remove(temp_log_file_name)
hardware_log[f"{source}_log"] = log
else:
raise RuntimeError(
f"`ConfigurationManager.download_log` did not create a `{source}`" f" file."
)
return hardware_log
