def __get_config_with_default_model_template_and_components(
self, config: Optional[ForecastConfig] = None) -> ForecastConfig:
"""Gets config with default value for `model_template` and `model_components_param` if not provided.
- model_template : default value is ``self.default_model_template_name``.
- model_components_param : default value is an empty ModelComponentsParam().
Parameters
----------
config : :class:`~greykite.framework.templates.model_templates.ForecastConfig` or None
Config object for template class to use.
See :class:`~greykite.framework.templates.model_templates.ForecastConfig`.
If None, uses an empty ForecastConfig.
Returns
-------
config : :class:`~greykite.framework.templates.model_templates.ForecastConfig`
Input ``config`` with default ``model_template`` populated.
If ``config.model_template`` is None, it is set to ``self.default_model_template_name``.
If ``config.model_components_param`` is None, it is set to ``ModelComponentsParam()``.
"""
config = config if config is not None else ForecastConfig()
# Unpacks list of a single element and sets default value if None.
# NB: Does not call `apply_forecast_config_defaults`.
# Only sets `model_template` and `model_components_param`.
# The template class may have its own implementation of forecast config defaults.
forecast_config_defaults = ForecastConfigDefaults()
forecast_config_defaults.DEFAULT_MODEL_TEMPLATE = self.default_model_template_name
config.model_template = forecast_config_defaults.apply_model_template_defaults(
config.model_template)
config.model_components_param = forecast_config_defaults.apply_model_components_defaults(
config.model_components_param)
return config
def test_apply_model_components_defaults():
"""Tests apply_model_components_defaults"""
assert ForecastConfigDefaults().apply_model_components_defaults(
None) == ModelComponentsParam()
mcp = ModelComponentsParam({"growth": "growth"})
assert ForecastConfigDefaults().apply_model_components_defaults(mcp) == mcp
assert ForecastConfigDefaults().apply_model_components_defaults([mcp
]) == mcp
assert ForecastConfigDefaults().apply_model_components_defaults(
[None, mcp]) == [ModelComponentsParam(), mcp]
def test_apply_model_template_defaults():
"""Tests apply_model_template_defaults"""
assert ForecastConfigDefaults().apply_model_template_defaults(
model_template=None) == "SILVERKITE"
mt = "RANDOM_TEMPLATE"
assert ForecastConfigDefaults().apply_model_template_defaults(
model_template=mt) == mt
assert ForecastConfigDefaults().apply_model_template_defaults(
model_template=[mt]) == mt
assert ForecastConfigDefaults().apply_model_template_defaults(
model_template=[None, mt]) == ["SILVERKITE", mt]
def assert_forecast_config_json_multiple_model_componments_parameter(
config: Optional[ForecastConfig] = None):
"""Asserts the forecast config values. This function expects a particular config and is not generic"""
config = ForecastConfigDefaults().apply_forecast_config_defaults(config)
assert config.model_template == [
ModelTemplateEnum.SILVERKITE.name,
ModelTemplateEnum.SILVERKITE_DAILY_90.name,
ModelTemplateEnum.SILVERKITE_WEEKLY.name
]
assert config.evaluation_metric_param.relative_error_tolerance == 0.02
# First model_components_param
model_components_param_1 = config.model_components_param[0]
assert model_components_param_1.autoregression is None
assert model_components_param_1.changepoints is None
assert model_components_param_1.custom is None
assert model_components_param_1.growth == {"growth_param": 0}
assert model_components_param_1.events == {"events_param": 1}
assert (model_components_param_1.hyperparameter_override is None
or model_components_param_1.hyperparameter_override is None)
assert model_components_param_1.regressors == {
"names": ["regressor1", "regressor2"]
}
assert model_components_param_1.lagged_regressors is None
assert model_components_param_1.seasonality == {"seas_param": 2}
assert model_components_param_1.uncertainty == {"uncertainty_param": 3}
# Second model_components_param
model_components_param_2 = config.model_components_param[1]
assert model_components_param_2.autoregression == {
"autoreg_dict": {
"autoreg_param": 0
}
}
assert model_components_param_2.changepoints is None
assert model_components_param_2.custom == {"custom_param": 1}
assert model_components_param_2.growth == {"growth_param": 2}
assert model_components_param_2.events == {"events_param": 3}
assert (model_components_param_2.hyperparameter_override == [{
"h1": 4
}, {
"h2": 5
}, None] or model_components_param_2.hyperparameter_override == [{
"h1": 4
}, {
"h2": 5
}, {}])
assert model_components_param_2.regressors == {
"names": ["regressor1", "regressor2"]
}
assert model_components_param_2.lagged_regressors == {
"lagged_regressor_dict": {
"lag_reg_param": 0
}
}
assert model_components_param_2.seasonality == {"seas_param": 6}
assert model_components_param_2.uncertainty == {"uncertainty_param": 7}
assert config.to_dict() # runs without error
def assert_default_forecast_config(config: Optional[ForecastConfig] = None):
"""Asserts for the default ForecastConfig values"""
try:
config = ForecastConfigDefaults().apply_forecast_config_defaults(
config)
assert config.model_template == ModelTemplateEnum.SILVERKITE.name
assert config.metadata_param.time_col == TIME_COL
assert config.metadata_param.value_col == VALUE_COL
assert config.evaluation_period_param.periods_between_train_test is None
assert config.evaluation_period_param.cv_max_splits == EVALUATION_PERIOD_CV_MAX_SPLITS
assert config.evaluation_metric_param.cv_selection_metric == EvaluationMetricEnum.MeanAbsolutePercentError.name
assert config.evaluation_metric_param.cv_report_metrics == CV_REPORT_METRICS_ALL
assert config.computation_param.n_jobs == COMPUTATION_N_JOBS
assert config.computation_param.verbose == COMPUTATION_VERBOSE
assert config.to_dict() # runs without error
except Exception:
fail("Config should not raise Exception")
def assert_forecast_config(config: Optional[ForecastConfig] = None):
"""Asserts the forecast config values. This function expects a particular config and is not generic"""
config = ForecastConfigDefaults().apply_forecast_config_defaults(config)
assert config.model_template == ModelTemplateEnum.SILVERKITE.name
assert config.metadata_param.time_col == "custom_time_col"
assert config.metadata_param.value_col == VALUE_COL
assert config.metadata_param.freq is None
assert config.metadata_param.date_format is None
assert config.metadata_param.train_end_date is None
assert config.metadata_param.anomaly_info == [{
"key": "value"
}, {
"key2": "value2"
}]
assert config.evaluation_period_param.test_horizon == 10
assert config.evaluation_period_param.periods_between_train_test == 5
assert config.evaluation_period_param.cv_horizon is None
assert config.evaluation_period_param.cv_min_train_periods == 20
assert config.evaluation_period_param.cv_expanding_window is True
assert config.evaluation_period_param.cv_use_most_recent_splits is None
assert config.evaluation_period_param.cv_periods_between_splits is None
assert config.evaluation_period_param.cv_periods_between_train_test == config.evaluation_period_param.periods_between_train_test
assert config.evaluation_period_param.cv_max_splits == EVALUATION_PERIOD_CV_MAX_SPLITS
assert config.evaluation_metric_param.cv_selection_metric == EvaluationMetricEnum.MeanSquaredError.name
assert config.evaluation_metric_param.cv_report_metrics == [
EvaluationMetricEnum.MeanAbsoluteError.name,
EvaluationMetricEnum.MeanAbsolutePercentError.name
]
assert config.evaluation_metric_param.agg_periods is None
assert config.evaluation_metric_param.agg_func is None
assert config.evaluation_metric_param.null_model_params is None
assert config.evaluation_metric_param.relative_error_tolerance == 0.02
assert config.model_components_param.autoregression == {
"autoreg_dict": {
"autoreg_param": 0
}
}
assert config.model_components_param.changepoints is None
assert config.model_components_param.custom == {"custom_param": 1}
assert config.model_components_param.growth == {"growth_param": 2}
assert config.model_components_param.events == {"events_param": 3}
assert (config.model_components_param.hyperparameter_override == [{
"h1": 4
}, {
"h2": 5
}, None] or config.model_components_param.hyperparameter_override
== [{
"h1": 4
}, {
"h2": 5
}, {}])
assert config.model_components_param.regressors == {
"names": ["regressor1", "regressor2"]
}
assert config.model_components_param.seasonality == {"seas_param": 6}
assert config.model_components_param.uncertainty == {
"uncertainty_param": 7
}
assert config.computation_param.hyperparameter_budget is None
assert config.computation_param.n_jobs == COMPUTATION_N_JOBS
assert config.computation_param.verbose == COMPUTATION_VERBOSE
assert config.to_dict() # runs without error