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

feat: Allow users to specify timestamp split for vertex forecasting #1187

Merged
merged 2 commits into from
May 3, 2022
Merged

feat: Allow users to specify timestamp split for vertex forecasting #1187

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
2 commits
Select commit Hold shift + click to select a range
68a1eb4
feat: Allow forecasting timestamp split.
TheMichaelHu Apr 27, 2022
a559434
Merge branch 'main' into mh-enable-timestamp-split
TheMichaelHu Apr 29, 2022
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
39 changes: 32 additions & 7 deletions google/cloud/aiplatform/training_jobs.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -408,7 +408,8 @@ def _create_input_data_config(
that piece is ignored by the pipeline.

Supported only for tabular and time series Datasets.
This parameter must be used with training_fraction_split, validation_fraction_split and test_fraction_split.
This parameter must be used with training_fraction_split,
validation_fraction_split, and test_fraction_split.
gcs_destination_uri_prefix (str):
Optional. The Google Cloud Storage location.

Expand Down Expand Up @@ -669,7 +670,8 @@ def _run_job(
that piece is ignored by the pipeline.

Supported only for tabular and time series Datasets.
This parameter must be used with training_fraction_split, validation_fraction_split and test_fraction_split.
This parameter must be used with training_fraction_split,
validation_fraction_split, and test_fraction_split.
model (~.model.Model):
Optional. Describes the Model that may be uploaded (via
[ModelService.UploadMode][]) by this TrainingPipeline. The
Expand Down Expand Up @@ -3487,9 +3489,9 @@ def run(
`time-offset` = `"Z"` (e.g. 1985-04-12T23:20:50.52Z). If for a
piece of data the key is not present or has an invalid value,
that piece is ignored by the pipeline.

Supported only for tabular and time series Datasets.
This parameter must be used with training_fraction_split, validation_fraction_split and test_fraction_split.
This parameter must be used with training_fraction_split,
validation_fraction_split, and test_fraction_split.
weight_column (str):
Optional. Name of the column that should be used as the weight column.
Higher values in this column give more importance to the row
Expand Down Expand Up @@ -3681,9 +3683,9 @@ def _run(
`time-offset` = `"Z"` (e.g. 1985-04-12T23:20:50.52Z). If for a
piece of data the key is not present or has an invalid value,
that piece is ignored by the pipeline.

Supported only for tabular and time series Datasets.
This parameter must be used with training_fraction_split, validation_fraction_split and test_fraction_split.
This parameter must be used with training_fraction_split,
validation_fraction_split, and test_fraction_split.
weight_column (str):
Optional. Name of the column that should be used as the weight column.
Higher values in this column give more importance to the row
Expand Down Expand Up @@ -4022,6 +4024,7 @@ def run(
validation_fraction_split: Optional[float] = None,
test_fraction_split: Optional[float] = None,
predefined_split_column_name: Optional[str] = None,
timestamp_split_column_name: Optional[str] = None,
weight_column: Optional[str] = None,
time_series_attribute_columns: Optional[List[str]] = None,
context_window: Optional[int] = None,
Expand Down Expand Up @@ -4106,6 +4109,16 @@ def run(
ignored by the pipeline.

Supported only for tabular and time series Datasets.
timestamp_split_column_name (str):
Optional. The key is a name of one of the Dataset's data
columns. The value of the key values of the key (the values in
the column) must be in RFC 3339 `date-time` format, where
`time-offset` = `"Z"` (e.g. 1985-04-12T23:20:50.52Z). If for a
piece of data the key is not present or has an invalid value,
that piece is ignored by the pipeline.
Supported only for tabular and time series Datasets.
This parameter must be used with training_fraction_split,
validation_fraction_split, and test_fraction_split.
weight_column (str):
Optional. Name of the column that should be used as the weight column.
Higher values in this column give more importance to the row
Expand Down Expand Up @@ -4229,6 +4242,7 @@ def run(
validation_fraction_split=validation_fraction_split,
test_fraction_split=test_fraction_split,
predefined_split_column_name=predefined_split_column_name,
timestamp_split_column_name=timestamp_split_column_name,
weight_column=weight_column,
time_series_attribute_columns=time_series_attribute_columns,
context_window=context_window,
Expand Down Expand Up @@ -4260,6 +4274,7 @@ def _run(
validation_fraction_split: Optional[float] = None,
test_fraction_split: Optional[float] = None,
predefined_split_column_name: Optional[str] = None,
timestamp_split_column_name: Optional[str] = None,
weight_column: Optional[str] = None,
time_series_attribute_columns: Optional[List[str]] = None,
context_window: Optional[int] = None,
Expand Down Expand Up @@ -4352,6 +4367,16 @@ def _run(
ignored by the pipeline.

Supported only for tabular and time series Datasets.
timestamp_split_column_name (str):
Optional. The key is a name of one of the Dataset's data
columns. The value of the key values of the key (the values in
the column) must be in RFC 3339 `date-time` format, where
`time-offset` = `"Z"` (e.g. 1985-04-12T23:20:50.52Z). If for a
piece of data the key is not present or has an invalid value,
that piece is ignored by the pipeline.
Supported only for tabular and time series Datasets.
This parameter must be used with training_fraction_split,
validation_fraction_split, and test_fraction_split.
weight_column (str):
Optional. Name of the column that should be used as the weight column.
Higher values in this column give more importance to the row
Expand Down Expand Up @@ -4511,7 +4536,7 @@ def _run(
validation_fraction_split=validation_fraction_split,
test_fraction_split=test_fraction_split,
predefined_split_column_name=predefined_split_column_name,
timestamp_split_column_name=None, # Not supported by AutoMLForecasting
timestamp_split_column_name=timestamp_split_column_name,
model=model,
create_request_timeout=create_request_timeout,
)
Expand Down
91 changes: 91 additions & 0 deletions tests/unit/aiplatform/test_automl_forecasting_training_jobs.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -148,6 +148,7 @@
_TEST_FRACTION_SPLIT_TEST = 0.2

_TEST_SPLIT_PREDEFINED_COLUMN_NAME = "split"
_TEST_SPLIT_TIMESTAMP_COLUMN_NAME = "timestamp"


@pytest.fixture
Expand Down Expand Up @@ -768,6 +769,96 @@ def test_splits_fraction(
timeout=None,
)

@pytest.mark.parametrize("sync", [True, False])
def test_splits_timestamp(
self,
mock_pipeline_service_create,
mock_pipeline_service_get,
mock_dataset_time_series,
mock_model_service_get,
sync,
):
"""Initiate aiplatform with encryption key name.

Create and run an AutoML Forecasting training job, verify calls and
return value
"""

aiplatform.init(
project=_TEST_PROJECT,
encryption_spec_key_name=_TEST_DEFAULT_ENCRYPTION_KEY_NAME,
)

job = AutoMLForecastingTrainingJob(
display_name=_TEST_DISPLAY_NAME,
optimization_objective=_TEST_TRAINING_OPTIMIZATION_OBJECTIVE_NAME,
column_transformations=_TEST_TRAINING_COLUMN_TRANSFORMATIONS,
)

model_from_job = job.run(
dataset=mock_dataset_time_series,
training_fraction_split=_TEST_FRACTION_SPLIT_TRAINING,
validation_fraction_split=_TEST_FRACTION_SPLIT_VALIDATION,
test_fraction_split=_TEST_FRACTION_SPLIT_TEST,
timestamp_split_column_name=_TEST_SPLIT_TIMESTAMP_COLUMN_NAME,
target_column=_TEST_TRAINING_TARGET_COLUMN,
time_column=_TEST_TRAINING_TIME_COLUMN,
time_series_identifier_column=_TEST_TRAINING_TIME_SERIES_IDENTIFIER_COLUMN,
unavailable_at_forecast_columns=_TEST_TRAINING_UNAVAILABLE_AT_FORECAST_COLUMNS,
available_at_forecast_columns=_TEST_TRAINING_AVAILABLE_AT_FORECAST_COLUMNS,
forecast_horizon=_TEST_TRAINING_FORECAST_HORIZON,
data_granularity_unit=_TEST_TRAINING_DATA_GRANULARITY_UNIT,
data_granularity_count=_TEST_TRAINING_DATA_GRANULARITY_COUNT,
model_display_name=_TEST_MODEL_DISPLAY_NAME,
weight_column=_TEST_TRAINING_WEIGHT_COLUMN,
time_series_attribute_columns=_TEST_TRAINING_TIME_SERIES_ATTRIBUTE_COLUMNS,
context_window=_TEST_TRAINING_CONTEXT_WINDOW,
budget_milli_node_hours=_TEST_TRAINING_BUDGET_MILLI_NODE_HOURS,
export_evaluated_data_items=_TEST_TRAINING_EXPORT_EVALUATED_DATA_ITEMS,
export_evaluated_data_items_bigquery_destination_uri=_TEST_TRAINING_EXPORT_EVALUATED_DATA_ITEMS_BIGQUERY_DESTINATION_URI,
export_evaluated_data_items_override_destination=_TEST_TRAINING_EXPORT_EVALUATED_DATA_ITEMS_OVERRIDE_DESTINATION,
quantiles=_TEST_TRAINING_QUANTILES,
validation_options=_TEST_TRAINING_VALIDATION_OPTIONS,
sync=sync,
create_request_timeout=None,
)

if not sync:
model_from_job.wait()

true_split = gca_training_pipeline.TimestampSplit(
training_fraction=_TEST_FRACTION_SPLIT_TRAINING,
validation_fraction=_TEST_FRACTION_SPLIT_VALIDATION,
test_fraction=_TEST_FRACTION_SPLIT_TEST,
key=_TEST_SPLIT_TIMESTAMP_COLUMN_NAME,
)

true_managed_model = gca_model.Model(
display_name=_TEST_MODEL_DISPLAY_NAME,
encryption_spec=_TEST_DEFAULT_ENCRYPTION_SPEC,
)

true_input_data_config = gca_training_pipeline.InputDataConfig(
timestamp_split=true_split, dataset_id=mock_dataset_time_series.name
)

true_training_pipeline = gca_training_pipeline.TrainingPipeline(
display_name=_TEST_DISPLAY_NAME,
training_task_definition=(
schema.training_job.definition.automl_forecasting
),
training_task_inputs=_TEST_TRAINING_TASK_INPUTS,
model_to_upload=true_managed_model,
input_data_config=true_input_data_config,
encryption_spec=_TEST_DEFAULT_ENCRYPTION_SPEC,
)

mock_pipeline_service_create.assert_called_once_with(
parent=initializer.global_config.common_location_path(),
training_pipeline=true_training_pipeline,
timeout=None,
)

@pytest.mark.parametrize("sync", [True, False])
def test_splits_predefined(
self,
Expand Down