Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Jump to bottom

feauture: upgrade job.py to results v2 format by default and expose h… #626

Closed
wants to merge 2 commits into from
Closed

feauture: upgrade job.py to results v2 format by default and expose h… #626

Show file tree
Hide file tree
Changes from 1 commit
Commits
Show all changes
2 commits
Select commit Hold shift + click to select a range
7957240
feauture: upgrade job.py to results v2 format by default and expose h…
kikomiss Jul 22, 2024
e7407a3
chore: updated recordings
kikomiss Jul 22, 2024
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
200 changes: 198 additions & 2 deletions azure-quantum/azure/quantum/job/job.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -108,6 +108,8 @@ def get_results(self, timeout_secs: float = DEFAULT_TIMEOUT):

Raises :class:`RuntimeError` if job execution fails.

Raises :class:`ValueError` if job output is malformed or output format is not compatible.

Raises :class:`azure.quantum.job.JobFailedWithResultsError` if job execution fails,
but failure results could still be retrieved (e.g. for jobs submitted against "microsoft.dft" target).

Expand Down Expand Up @@ -142,21 +144,215 @@ def get_results(self, timeout_secs: float = DEFAULT_TIMEOUT):

if self.details.output_data_format == "microsoft.quantum-results.v1":
if "Histogram" not in results:
raise f"\"Histogram\" array was expected to be in the Job results for \"{self.details.output_data_format}\" output format."
raise ValueError(f"\"Histogram\" array was expected to be in the Job results for \"{self.details.output_data_format}\" output format.")

histogram_values = results["Histogram"]

if len(histogram_values) % 2 == 0:
# Re-mapping {'Histogram': ['[0]', 0.50, '[1]', 0.50] } to {'[0]': 0.50, '[1]': 0.50}
return {histogram_values[i]: histogram_values[i + 1] for i in range(0, len(histogram_values), 2)}
else:
raise f"\"Histogram\" array has invalid format. Even number of items is expected."
raise ValueError(f"\"Histogram\" array has invalid format. Even number of items is expected.")
elif self.details.output_data_format == "microsoft.quantum-results.v2":
if "DataFormat" not in results or results["DataFormat"] != "microsoft.quantum-results.v2":
raise ValueError(f"\"DataFormat\" was expected to be \"microsoft.quantum-results.v2\" in the Job results for \"{self.details.output_data_format}\" output format.")

if "Results" not in results:
raise ValueError(f"\"Results\" field was expected to be in the Job results for \"{self.details.output_data_format}\" output format.")

if len(results["Results"]) < 1:
raise ValueError("\"Results\" array was expected to contain at least one item")

results = results["Results"][0]

if "Histogram" not in results:
raise ValueError(f"\"Histogram\" array was expected to be in the Job results for \"{self.details.output_data_format}\" output format.")

if "Shots" not in results:
raise ValueError(f"\"Shots\" array was expected to be in the Job results for \"{self.details.output_data_format}\" output format.")

histogram_values = results["Histogram"]

total_count = len(results["Shots"])

# Re-mapping object {'Histogram': [{"Outcome": [0], "Display": '[0]', "Count": 500}, {"Outcome": [1], "Display": '[1]', "Count": 500}]} to {'[0]': 0.50, '[1]': 0.50}
return {outcome["Display"]: outcome["Count"] / total_count for outcome in histogram_values}

return results
except:
# If errors decoding the data, return the raw payload:
return payload

def get_results_histogram(self, timeout_secs: float = DEFAULT_TIMEOUT):
"""Get job results histogram by downloading the results blob from the storage container linked via the workspace.

Raises :class:`RuntimeError` if job execution fails.

Raises :class:`ValueError` if job output is malformed or output format is not compatible.

Raises :class:`azure.quantum.job.JobFailedWithResultsError` if job execution fails,
but failure results could still be retrieved (e.g. for jobs submitted against "microsoft.dft" target).

:param timeout_secs: Timeout in seconds, defaults to 300
:type timeout_secs: float
:return: Results dictionary with histogram shots, or raw results if not a json object.
:rtype: typing.Any
"""
if self.results is not None:
return self.results

if not self.has_completed():
self.wait_until_completed(timeout_secs=timeout_secs)

if not self.details.status == "Succeeded":
if self.details.status == "Failed" and self._allow_failure_results():
job_blob_properties = self.download_blob_properties(self.details.output_data_uri)
if job_blob_properties.size > 0:
job_failure_data = self.download_data(self.details.output_data_uri)
raise JobFailedWithResultsError("An error occurred during job execution.", job_failure_data)

raise RuntimeError(
f'{"Cannot retrieve results as job execution failed"}'
+ f"(status: {self.details.status}."
+ f"error: {self.details.error_data})"
)

payload = self.download_data(self.details.output_data_uri)
try:
payload = payload.decode("utf8")
results = json.loads(payload)

if self.details.output_data_format == "microsoft.quantum-results.v2":
if "DataFormat" not in results or results["DataFormat"] != "microsoft.quantum-results.v2":
raise ValueError(f"\"DataFormat\" was expected to be \"microsoft.quantum-results.v2\" in the Job results for \"{self.details.output_data_format}\" output format.")
if "Results" not in results:
raise ValueError(f"\"Results\" field was expected to be in the Job results for \"{self.details.output_data_format}\" output format.")

if len(results["Results"]) < 1:
raise ValueError("\"Results\" array was expected to contain at least one item")

results = results["Results"]

if len(results) == 1:
results = results[0]
if "Histogram" not in results:
raise ValueError(f"\"Histogram\" array was expected to be in the Job results for \"{self.details.output_data_format}\" output format.")

histogram_values = results["Histogram"]
outcome_keys = self._process_outcome(histogram_values)

# Re-mapping object {'Histogram': [{"Outcome": [0], "Display": '[0]', "Count": 500}, {"Outcome": [1], "Display": '[1]', "Count": 500}]} to {'[0]': {"Outcome": [0], "Count": 500}, '[1]': {"Outcome": [1], "Count": 500}}
return {hist_val["Display"]: {"outcome": outcome, "count": hist_val["Count"]} for outcome, hist_val in zip(outcome_keys, histogram_values)}

else:
# This is handling the BatchResults edge case
resultsArray = []
for i, result in enumerate(results):
if "Histogram" not in result:
raise ValueError(f"\"Histogram\" array was expected to be in the Job results for result {i} for \"{self.details.output_data_format}\" output format.")

histogram_values = result["Histogram"]
outcome_keys = self._process_outcome(histogram_values)

# Re-mapping object {'Histogram': [{"Outcome": [0], "Display": '[0]', "Count": 500}, {"Outcome": [1], "Display": '[1]', "Count": 500}]} to {'[0]': {"Outcome": [0], "Count": 500}, '[1]': {"Outcome": [1], "Count": 500}}
resultsArray.append({hist_val["Display"]: {"outcome": outcome, "count": hist_val["Count"]} for outcome, hist_val in zip(outcome_keys, histogram_values)})

return resultsArray

else:
raise ValueError(f"Getting a results histogram with counts instead of probabilities is not a supported feature for jobs using the \"{self.details.output_data_format}\" output format.")

except Exception as e:
raise e

def get_results_shots(self, timeout_secs: float = DEFAULT_TIMEOUT):
"""Get job results per shot data by downloading the results blob from the
storage container linked via the workspace.

Raises :class:`RuntimeError` if job execution fails.

Raises :class:`ValueError` if job output is malformed or output format is not compatible.

Raises :class:`azure.quantum.job.JobFailedWithResultsError` if job execution fails,
but failure results could still be retrieved (e.g. for jobs submitted against "microsoft.dft" target).

:param timeout_secs: Timeout in seconds, defaults to 300
:type timeout_secs: float
:return: Results dictionary with histogram shots, or raw results if not a json object.
:rtype: typing.Any
"""
if self.results is not None:
return self.results

if not self.has_completed():
self.wait_until_completed(timeout_secs=timeout_secs)

if not self.details.status == "Succeeded":
if self.details.status == "Failed" and self._allow_failure_results():
job_blob_properties = self.download_blob_properties(self.details.output_data_uri)
if job_blob_properties.size > 0:
job_failure_data = self.download_data(self.details.output_data_uri)
raise JobFailedWithResultsError("An error occurred during job execution.", job_failure_data)

raise RuntimeError(
f'{"Cannot retrieve results as job execution failed"}'
+ f"(status: {self.details.status}."
+ f"error: {self.details.error_data})"
)

payload = self.download_data(self.details.output_data_uri)
try:
payload = payload.decode("utf8")
results = json.loads(payload)

if self.details.output_data_format == "microsoft.quantum-results.v2":
if "DataFormat" not in results or results["DataFormat"] != "microsoft.quantum-results.v2":
raise ValueError(f"\"DataFormat\" was expected to be \"microsoft.quantum-results.v2\" in the Job results for \"{self.details.output_data_format}\" output format.")
if "Results" not in results:
raise ValueError(f"\"Results\" field was expected to be in the Job results for \"{self.details.output_data_format}\" output format.")

results = results["Results"]

if len(results) < 1:
raise ValueError("\"Results\" array was expected to contain at least one item")

if len(results) == 1:
result = results[0]
if "Shots" not in result:
raise ValueError(f"\"Shots\" array was expected to be in the Job results for \"{self.details.output_data_format}\" output format.")

return [self._convert_tuples(shot) for shot in result["Shots"]]
else:
# This is handling the BatchResults edge case
shotsArray = []
for i, result in enumerate(results):
if "Shots" not in result:
raise ValueError(f"\"Shots\" array was expected to be in the Job results for result {i} of \"{self.details.output_data_format}\" output format.")
shotsArray.append([self._convert_tuples(shot) for shot in result["Shots"]])

return shotsArray
else:
raise ValueError(f"Individual shot results are not supported for jobs using the \"{self.details.output_data_format}\" output format.")
except Exception as e:
raise e

def _process_outcome(self, histogram_results):
return [self._convert_tuples(v['Outcome']) for v in histogram_results]

def _convert_tuples(self, data):
if isinstance(data, dict):
# Check if the dictionary represents a tuple
if all(isinstance(k, str) and k.startswith("Item") for k in data.keys()):
# Convert the dictionary to a tuple
return tuple(self._convert_tuples(data[f"Item{i+1}"]) for i in range(len(data)))
else:
raise "Malformed tuple output"
elif isinstance(data, list):
# Recursively process list elements
return [self._convert_tuples(item) for item in data]
else:
# Return the data as is (int, string, etc.)
return data

@classmethod
def _allow_failure_results(cls) -> bool:
Expand Down
2 changes: 1 addition & 1 deletion azure-quantum/azure/quantum/target/target.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -189,7 +189,7 @@ def _encode_input_data(data: Any) -> bytes:

def _qir_output_data_format(self) -> str:
""""Fallback output data format in case of QIR job submission."""
return "microsoft.quantum-results.v1"
return "microsoft.quantum-results.v2"

def submit(
self,
Expand Down
Loading
Loading