Skip to content

chore: Handle interactions_data_source as a valid data source #7005

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

Merged
copybara-service merged 1 commit into from
Jul 21, 2026
+120 −0
Merged

chore: Handle interactions_data_source as a valid data source #7005

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
1 commit
Select commit
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
8 changes: 8 additions & 0 deletions agentplatform/_genai/_evals_data_converters.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -796,6 +796,10 @@ def merge_evaluation_datasets(
candidate_responses.append(base_eval_case.responses[0])
elif base_eval_case.agent_data:
candidate_responses.append(_create_placeholder_response_candidate(""))
elif getattr(base_eval_case, "interactions_data_source", None):
# Interaction data will be resolved server-side for metric
# computation; add a placeholder without warning.
candidate_responses.append(_create_placeholder_response_candidate(""))
else:
logger.warning(
"No response or agent data found for base dataset (index 0) in case %s. "
Expand Down Expand Up @@ -853,6 +857,10 @@ def merge_evaluation_datasets(
candidate_responses.append(current_ds_eval_case.responses[0])
elif current_ds_eval_case.agent_data:
candidate_responses.append(_create_placeholder_response_candidate(""))
elif getattr(current_ds_eval_case, "interactions_data_source", None):
# Interaction data will be resolved server-side for metric
# computation; add a placeholder without warning.
candidate_responses.append(_create_placeholder_response_candidate(""))
else:
logger.warning(
"No response or agent data found for dataset %s in case %s. Adding"
Expand Down
112 changes: 112 additions & 0 deletions tests/unit/agentplatform/genai/test_evals.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -7752,6 +7752,118 @@ def test_merge_with_invalid_eval_case_type(self):
schemas=[_evals_data_converters.EvalDatasetSchema.FLATTEN],
)

def test_merge_single_dataset_with_interactions_data_source(self):
"""Base dataset with interactions_data_source adds placeholder without warning."""
dataset = agentplatform_genai_types.EvaluationDataset(
eval_cases=[
agentplatform_genai_types.EvalCase(
interactions_data_source=agentplatform_genai_types.InteractionsDataSource(
interaction="projects/p/locations/l/interactions/i1",
),
),
]
)

with mock.patch.object(
_evals_data_converters, "logger"
) as mock_logger:
merged = _evals_data_converters.merge_evaluation_datasets([dataset])

assert len(merged.eval_cases) == 1
assert len(merged.eval_cases[0].responses) == 1
assert merged.eval_cases[0].responses[0].response == genai_types.Content(
parts=[genai_types.Part(text="")]
)
mock_logger.warning.assert_not_called()

def test_merge_two_datasets_with_interactions_data_source(self):
"""Merging two interaction-id datasets adds placeholders without warning."""
dataset_1 = agentplatform_genai_types.EvaluationDataset(
eval_cases=[
agentplatform_genai_types.EvalCase(
interactions_data_source=agentplatform_genai_types.InteractionsDataSource(
interaction="projects/p/locations/l/interactions/i1",
),
),
]
)
dataset_2 = agentplatform_genai_types.EvaluationDataset(
eval_cases=[
agentplatform_genai_types.EvalCase(
interactions_data_source=agentplatform_genai_types.InteractionsDataSource(
interaction="projects/p/locations/l/interactions/i2",
),
),
]
)

with mock.patch.object(
_evals_data_converters, "logger"
) as mock_logger:
merged = _evals_data_converters.merge_evaluation_datasets(
[dataset_1, dataset_2]
)

assert len(merged.eval_cases) == 1
assert len(merged.eval_cases[0].responses) == 2
assert merged.eval_cases[0].responses[0].response == genai_types.Content(
parts=[genai_types.Part(text="")]
)
assert merged.eval_cases[0].responses[1].response == genai_types.Content(
parts=[genai_types.Part(text="")]
)
mock_logger.warning.assert_not_called()

def test_merge_interactions_data_source_with_response_dataset(self):
"""Merging an interaction-id dataset with a response dataset works correctly."""
dataset_interactions = agentplatform_genai_types.EvaluationDataset(
eval_cases=[
agentplatform_genai_types.EvalCase(
prompt=genai_types.Content(
parts=[genai_types.Part(text="Prompt 1")]
),
interactions_data_source=agentplatform_genai_types.InteractionsDataSource(
interaction="projects/p/locations/l/interactions/i1",
),
),
]
)
dataset_response = agentplatform_genai_types.EvaluationDataset(
eval_cases=[
agentplatform_genai_types.EvalCase(
prompt=genai_types.Content(
parts=[genai_types.Part(text="Prompt 1")]
),
responses=[
agentplatform_genai_types.ResponseCandidate(
response=genai_types.Content(
parts=[genai_types.Part(text="Response 1")]
)
)
],
),
]
)

with mock.patch.object(
_evals_data_converters, "logger"
) as mock_logger:
merged = _evals_data_converters.merge_evaluation_datasets(
[dataset_interactions, dataset_response]
)

assert len(merged.eval_cases) == 1
assert len(merged.eval_cases[0].responses) == 2
# First response is a placeholder from the interactions_data_source case
assert merged.eval_cases[0].responses[0].response == genai_types.Content(
parts=[genai_types.Part(text="")]
)
# Second response is the actual response
assert merged.eval_cases[0].responses[1].response == genai_types.Content(
parts=[genai_types.Part(text="Response 1")]
)
mock_logger.warning.assert_not_called()


@pytest.mark.usefixtures("google_auth_mock")
class TestPredefinedMetricHandler:
Expand Down
Loading