Batch

`Batch(**data)`

Bases: BaseModel

Class to load batch data return by the API.

A batch groups up several jobs with the same sequence. When
a batch is assigned to a QPU, all its jobs are run sequentially
and no other batch can be assigned to the device until all its
jobs are done and declared complete.

ATTRIBUTE	DESCRIPTION
`open`	Whether the batch accepts more jobs or not. TYPE: `bool`
`complete`	Opposite of open, deprecated. TYPE: `bool`
`created_at`	Timestamp of the creation of the batch. TYPE: `str`
`updated_at`	Timestamp of the last update of the batch. TYPE: `str`
`device_type`	Type of device to run the batch on. TYPE: `str`
`project_id`	ID of the owner project of the batch. TYPE: `str`
`id`	Unique identifier for the batch. TYPE: `str`
`user_id`	Unique identifier of the user that created the batch. TYPE: `str`
`status`	Status of the batch. Possible values are: PENDING, RUNNING, DONE, CANCELED, TIMED_OUT, ERROR, PAUSED. TYPE: `str`
`_client`	A Client instance to connect to PCS. TYPE: `Client`
`ordered_jobs`	List of all the jobs added to the batch, ordered by creation time. TYPE: `List[Job]`
`jobs_count`	Number of jobs added to the batch. TYPE: `int`
`jobs_count_per_status`	Number of jobs per status. TYPE: `Dict[str, int]`
`webhook`	Webhook where the job results are automatically sent to. TYPE: `Optional[str]`
`start_datetime`	Timestamp of the time the batch was sent to the QPU. TYPE: `Optional[str]`
`end_datetime`	Timestamp of when the batch process was finished. TYPE: `Optional[str]`
`device_status`	Status of the device where the batch is running. TYPE: `Optional[str]`
`parent_id`	ID from previous batch if a batch was resubmitted for execution. TYPE: `Optional[str]`
`configuration`	Further configuration for certain emulators. TYPE: `Union[BaseConfig, Dict[str, Any], None]`
`backend_configuration`	serialised config object for emulation runtimes. This is a configuration from the pulser library that can be used on certain backend emulator types. It must be in the form of an abstract representation/encoded string. TYPE: `Optional[str]`
`tags`	Keyword used to refine the batch search. TYPE: `Optional[list[str]]`
`jobs`	Dictionary of all the jobs added to the batch. TYPE: `deprecated`
`sequence_builder`	Pulser sequence of the batch. TYPE: `Optional[str]`

Source code in pasqal_cloud/batch.py

def __init__(self, **data: Any) -> None:
    # Workaround to make the private attribute '_client' working
    # like we need with Pydantic V2, more information on
    # https://docs.pydantic.dev/latest/concepts/models/#private-model-attributes
    super().__init__(**data)
    self._client = data["_client"]
    if data.get("jobs"):
        self._ordered_jobs = [
            Job(**raw_job, _client=self._client) for raw_job in data["jobs"]
        ]

`jobs` `property` `writable`

Once the 'ordered_jobs' is built, we need to keep the 'jobs' attribute for backward compatibility with the code written by the users of the sdk

`add_jobs(jobs, wait=False)`

Add some jobs to batch for execution on PCS and returns the updated batch.

The batch must be open otherwise the API will return an error. The new jobs are appended to the ordered_jobs list attribute.

PARAMETER	DESCRIPTION
`jobs`	List of jobs to be added to the batch. TYPE: `List[CreateJob]`
`wait`	If True, blocks until all jobs in the batch are done. TYPE: `bool` DEFAULT: `False`

Source code in pasqal_cloud/batch.py

def add_jobs(
    self,
    jobs: List[CreateJob],
    wait: bool = False,
) -> None:
    """Add some jobs to batch for execution on PCS and returns the updated batch.

    The batch must be `open` otherwise the API will return an error.
    The new jobs are appended to the `ordered_jobs` list attribute.

    Args:
        jobs: List of jobs to be added to the batch.
        wait: If True, blocks until all jobs in the batch are done.

    """
    try:
        batch_rsp = self._client.add_jobs(self.id, jobs)
    except HTTPError as e:
        raise JobCreationError(e) from e
    self._update_from_api_response(batch_rsp)

    if wait:
        while any(
            job.status in {"PENDING", "RUNNING"} for job in self.ordered_jobs
        ):
            time.sleep(RESULT_POLLING_INTERVAL)
            self.refresh()

`cancel()`

Cancel the current batch on the PCS.

Source code in pasqal_cloud/batch.py

def cancel(self) -> None:
    """Cancel the current batch on the PCS."""
    try:
        batch_rsp = self._client.cancel_batch(self.id)
    except HTTPError as e:
        raise BatchCancellingError(e) from e
    self._update_from_api_response(batch_rsp)

`close(wait=False, fetch_results=False)`

Declare to PCS that the batch is closed and returns an updated batch instance.

PARAMETER	DESCRIPTION
`wait`	Whether to wait for the batch to be done and fetch results. TYPE: `bool` DEFAULT: `False`
`fetch_results`	Whether to wait for the batch to be done and fetch results. TYPE: `deprecated` DEFAULT: `False`

A batch that is closed awaits no extra jobs. All jobs previously added will be executed before the batch is terminated. When all its jobs are done, the closed batch is unassigned to its running device.

Source code in pasqal_cloud/batch.py

def close(self, wait: bool = False, fetch_results: bool = False) -> None:
    """Declare to PCS that the batch is closed and returns an updated
    batch instance.

    Args:
        wait: Whether to wait for the batch to be done and fetch results.
        fetch_results (deprecated): Whether to wait for the batch \
            to be done and fetch results.

    A batch that is closed awaits no extra jobs. All jobs previously added
    will be executed before the batch is terminated. When all its jobs are done,
    the closed batch is unassigned to its running device.
    """
    try:
        batch_rsp = self._client.close_batch(self.id)
    except HTTPError as e:
        raise BatchClosingError(e) from e
    self._update_from_api_response(batch_rsp)
    if wait or fetch_results:
        while any(
            job.status in {"PENDING", "RUNNING"} for job in self.ordered_jobs
        ):
            time.sleep(RESULT_POLLING_INTERVAL)
            self.refresh()

`declare_complete(wait=False, fetch_results=False)`

Deprecated, use close instead.

Source code in pasqal_cloud/batch.py

def declare_complete(self, wait: bool = False, fetch_results: bool = False) -> None:
    """
    Deprecated, use close instead.
    """
    warn(
        "This method is deprecated, use close instead.",
        DeprecationWarning,
        stacklevel=2,
    )
    return self.close(wait=wait, fetch_results=fetch_results)

`refresh()`

Fetch the batch from the API and update it in place.

Source code in pasqal_cloud/batch.py

def refresh(self) -> None:
    """Fetch the batch from the API and update it in place."""
    try:
        batch_rsp = self._client.get_batch(self.id)
    except HTTPError as e:
        raise BatchFetchingError(e) from e
    self._update_from_api_response(batch_rsp)

`retry(job, wait=False)`

Retry a job in the same batch. The batch should not be 'closed'. The new job is appended to the ordered_jobs list attribute.

PARAMETER	DESCRIPTION
`job`	The job to retry TYPE: `Job`
`wait`	Whether to wait for job completion TYPE: `bool` DEFAULT: `False`

RAISES	DESCRIPTION
`JobRetryError`	if there was an error adding the job to the batch.

Source code in pasqal_cloud/batch.py

def retry(self, job: Job, wait: bool = False) -> None:
    """
    Retry a job in the same batch.
    The batch should not be 'closed'.
    The new job is appended to the `ordered_jobs` list attribute.

    Args:
        job: The job to retry
        wait: Whether to wait for job completion

    Raises:
        JobRetryError: if there was an error adding the job to the batch.
    """
    retried_job = CreateJob(runs=job.runs, variables=job.variables)
    try:
        self.add_jobs([retried_job], wait=wait)
    except JobCreationError as e:
        raise JobRetryError from e

`set_tags(tags)`

Set tags for the current batch, overwriting previous ones already set

Source code in pasqal_cloud/batch.py

def set_tags(
    self,
    tags: list[str],
) -> None:
    """Set tags for the current batch, overwriting previous ones already set"""
    try:
        batch_rsp = self._client.set_batch_tags(self.id, tags)
    except HTTPError as e:
        raise BatchSetTagsError(e) from e
    return self._update_from_api_response(batch_rsp)

Previous
Job Next
Workload

Batch

Batch(**data)

jobs property writable

add_jobs(jobs, wait=False)

cancel()

close(wait=False, fetch_results=False)

declare_complete(wait=False, fetch_results=False)

refresh()

retry(job, wait=False)

set_tags(tags)

`Batch(**data)`

`jobs` `property` `writable`

`add_jobs(jobs, wait=False)`

`cancel()`

`close(wait=False, fetch_results=False)`

`declare_complete(wait=False, fetch_results=False)`

`refresh()`

`retry(job, wait=False)`

`set_tags(tags)`