Swap is_multiseries logic to problem type (#4278)

* Add multiseries time series regression as problem type

* Completely revamp to multiseries based on problem type

evalml/automl/automl_algorithm/default_algorithm.py

@@ -25,7 +25,7 @@
     _make_pipeline_from_multiple_graphs,
     make_pipeline,
 )
-from evalml.problem_types import is_regression, is_time_series
+from evalml.problem_types import is_multiseries, is_regression, is_time_series
 from evalml.utils import infer_feature_types
 from evalml.utils.logger import get_logger
 
@@ -81,7 +81,6 @@ class DefaultAlgorithm(AutoMLAlgorithm):
             model families. Run evalml.pipelines.components.utils.allowed_model_families("binary") to see options. Change `binary`
             to `multiclass` or `regression` depending on the problem type.
         excluded_model_families (list[ModelFamily]): A list of model families to exclude from the estimators used when building pipelines. For default algorithm, this only excludes estimators in the non-naive batches.
-        is_multiseries (bool): Whether or not the problem is a multiseries time series problem. Defaults to False.
     """
 
     def __init__(
@@ -106,7 +105,6 @@ def __init__(
         run_feature_selection=True,
         verbose=False,
         exclude_featurizers=None,
-        is_multiseries=False,
     ):
         super().__init__(
             allowed_pipelines=[],
@@ -140,7 +138,6 @@ def __init__(
         self.run_feature_selection = run_feature_selection
         self.ensembling = ensembling
         self.exclude_featurizers = exclude_featurizers or []
-        self.is_multiseries = is_multiseries
 
         if allowed_model_families is not None and excluded_model_families is not None:
             raise ValueError(
@@ -173,7 +170,7 @@ def default_max_batches(self):
         """Returns the number of max batches AutoMLSearch should run by default."""
         if self.ensembling:
             return 3
-        elif self.is_multiseries:
+        elif is_multiseries(self.problem_type):
             return 1
         else:
             return 2
@@ -222,7 +219,6 @@ def _non_naive_estimators(self):
                 self.problem_type,
                 model_families=self.allowed_model_families,
                 excluded_model_families=self.excluded_model_families,
-                is_multiseries=self.is_multiseries,
             )
             if est not in self._naive_estimators()
         ]
@@ -271,7 +267,6 @@ def _create_naive_pipelines(self, use_features=False):
                 ),
                 features=self.features,
                 exclude_featurizers=self.exclude_featurizers,
-                is_multiseries=self.is_multiseries,
             )
             for estimator in estimators
         ]
@@ -300,7 +295,6 @@ def _add_without_pipelines(self, pipelines, estimators, feature_selector=[]):
                     features=self.features,
                     exclude_featurizers=self.exclude_featurizers,
                     include_decomposer=False,
-                    is_multiseries=self.is_multiseries,
                 )
                 for estimator in estimators
             ]
@@ -440,7 +434,6 @@ def _make_pipelines_helper(self, estimators):
                     ),
                     features=self.features,
                     exclude_featurizers=self.exclude_featurizers,
-                    is_multiseries=self.is_multiseries,
                 )
                 for estimator in estimators
             ]
@@ -484,7 +477,7 @@ def next_batch(self):
             # Skip the naive batch for multiseries time series
             batch = (
                 self._batch_number
-                if not self.is_multiseries
+                if not is_multiseries(self.problem_type)
                 else self._batch_number + 1
             )
             if batch == 0:
@@ -679,7 +672,6 @@ def _make_split_pipeline(self, estimator, pipeline_name=None):
                 extra_components_before=[SelectColumns],
                 use_estimator=False,
                 exclude_featurizers=self.exclude_featurizers,
-                is_multiseries=self.is_multiseries,
             )
 
             numeric_pipeline = make_pipeline(
@@ -693,7 +685,6 @@ def _make_split_pipeline(self, estimator, pipeline_name=None):
                 extra_components_after=[SelectColumns],
                 use_estimator=False,
                 exclude_featurizers=self.exclude_featurizers,
-                is_multiseries=self.is_multiseries,
             )
             pre_pipeline_components = (
                 {"DFS Transformer": ["DFS Transformer", "X", "y"]}
@@ -745,7 +736,6 @@ def _make_split_pipeline(self, estimator, pipeline_name=None):
                 extra_components_before=[SelectColumns],
                 features=self.features,
                 exclude_featurizers=self.exclude_featurizers,
-                is_multiseries=self.is_multiseries,
             )
             return categorical_pipeline
         elif self.run_feature_selection:
@@ -762,7 +752,6 @@ def _make_split_pipeline(self, estimator, pipeline_name=None):
                 extra_components_after=[SelectColumns],
                 features=self.features,
                 exclude_featurizers=self.exclude_featurizers,
-                is_multiseries=self.is_multiseries,
             )
             return numeric_pipeline
 
@@ -774,6 +763,5 @@ def _make_split_pipeline(self, estimator, pipeline_name=None):
                 self.problem_type,
                 sampler_name=self.sampler_name,
                 exclude_featurizers=self.exclude_featurizers,
-                is_multiseries=self.is_multiseries,
             )
             return pipeline

evalml/automl/automl_algorithm/iterative_algorithm.py

@@ -69,7 +69,6 @@ class IterativeAlgorithm(AutoMLAlgorithm):
         verbose (boolean): Whether or not to display logging information regarding pipeline building. Defaults to False.
         exclude_featurizers (list[str]): A list of featurizer components to exclude from the pipelines built by IterativeAlgorithm.
             Valid options are "DatetimeFeaturizer", "EmailFeaturizer", "URLFeaturizer", "NaturalLanguageFeaturizer", "TimeSeriesFeaturizer"
-        is_multiseries (bool): Whether or not the problem is a multiseries time series problem. Defaults to False.
     """
 
     def __init__(
@@ -96,7 +95,6 @@ def __init__(
         features=None,
         verbose=False,
         exclude_featurizers=None,
-        is_multiseries=False,
     ):
         self.X = infer_feature_types(X)
         self.y = infer_feature_types(y)
@@ -131,7 +129,6 @@ def __init__(
         self.features = features
         self._set_additional_pipeline_params()
         self.exclude_featurizers = exclude_featurizers
-        self.is_multiseries = is_multiseries
 
         super().__init__(
             allowed_pipelines=self.allowed_pipelines,
@@ -159,7 +156,6 @@ def _create_pipelines(self):
                 self.problem_type,
                 model_families=self.allowed_model_families,
                 excluded_model_families=self.excluded_model_families,
-                is_multiseries=self.is_multiseries,
             )
             allowed_estimators = self._filter_estimators(
                 allowed_estimators,
@@ -188,7 +184,6 @@ def _create_pipelines(self):
                         ).get("known_in_advance", None),
                         features=self.features,
                         exclude_featurizers=self.exclude_featurizers,
-                        is_multiseries=self.is_multiseries,
                     )
                     for estimator in allowed_estimators
                 ]
@@ -212,7 +207,6 @@ def _create_pipelines(self):
                             features=self.features,
                             exclude_featurizers=self.exclude_featurizers,
                             include_decomposer=False,
-                            is_multiseries=self.is_multiseries,
                         )
                         for estimator in allowed_estimators
                     ]

evalml/automl/automl_search.py

@@ -65,6 +65,7 @@
     handle_problem_types,
     is_binary,
     is_classification,
+    is_multiseries,
     is_time_series,
 )
 from evalml.tuners import SKOptTuner
@@ -625,10 +626,6 @@ def __init__(
         self.problem_configuration = self._validate_problem_configuration(
             problem_configuration,
         )
-        self.is_multiseries = (
-            is_time_series(self.problem_type)
-            and self.problem_configuration.get("series_id") is not None
-        )
         self._train_best_pipeline = train_best_pipeline
         self._best_pipeline = None
         self._searched = False
@@ -657,7 +654,7 @@ def __init__(
                 )
 
         # For multiseries problems, we need to mke sure that the data is primarily ordered by the time_index rather than the series_id
-        if self.is_multiseries:
+        if is_multiseries(self.problem_type):
             time_index = self.problem_configuration.get("time_index")
             series_id = self.problem_configuration.get("series_id")
             X_train = X_train.sort_values([time_index, series_id])
@@ -946,7 +943,6 @@ def _is_imbalanced(X, y, problem_type):
                 features=features,
                 verbose=self.verbose,
                 exclude_featurizers=self.exclude_featurizers,
-                is_multiseries=self.is_multiseries,
             )
         elif automl_algorithm == "default":
             self.automl_algorithm = DefaultAlgorithm(
@@ -967,7 +963,6 @@ def _is_imbalanced(X, y, problem_type):
                 verbose=self.verbose,
                 n_jobs=self.n_jobs,
                 exclude_featurizers=self.exclude_featurizers,
-                is_multiseries=self.is_multiseries,
             )
         else:
             raise ValueError("Please specify a valid automl algorithm.")
@@ -1068,6 +1063,13 @@ def _validate_problem_configuration(self, problem_configuration=None):
             is_valid, msg = contains_all_ts_parameters(problem_configuration)
             if not is_valid:
                 raise ValueError(msg)
+            if (
+                is_multiseries(self.problem_type)
+                and "series_id" not in problem_configuration
+            ):
+                raise ValueError(
+                    "Must provide 'series_id' column in problem_configuration for multiseries time series problems.",
+                )
         return problem_configuration or {}
 
     def _handle_keyboard_interrupt(self):
@@ -1380,7 +1382,6 @@ def _get_baseline_pipeline(self):
                 forecast_horizon,
                 time_index,
                 exclude_timeseries_featurizer,
-                self.is_multiseries,
                 series_id,
             )
         return baseline

evalml/automl/utils.py

@@ -49,6 +49,7 @@ def get_default_primary_search_objective(problem_type):
         "time series regression": "MedianAE",
         "time series binary": "Log Loss Binary",
         "time series multiclass": "Log Loss Multiclass",
+        "multiseries time series regression": "MedianAE",
     }[problem_type.value]
     return get_objective(objective_name, return_instance=True)
 

evalml/objectives/regression_objective.py

@@ -6,5 +6,9 @@
 class RegressionObjective(ObjectiveBase):
     """Base class for all regression objectives."""
 
-    problem_types = [ProblemTypes.REGRESSION, ProblemTypes.TIME_SERIES_REGRESSION]
-    """[ProblemTypes.REGRESSION, ProblemTypes.TIME_SERIES_REGRESSION]"""
+    problem_types = [
+        ProblemTypes.REGRESSION,
+        ProblemTypes.TIME_SERIES_REGRESSION,
+        ProblemTypes.MULTISERIES_TIME_SERIES_REGRESSION,
+    ]
+    """[ProblemTypes.REGRESSION, ProblemTypes.TIME_SERIES_REGRESSION, ProblemTypes.MULTISERIES_TIME_SERIES_REGRESSION]"""

evalml/pipelines/components/component_base.py

@@ -31,7 +31,6 @@ class ComponentBase(ABC, metaclass=ComponentBaseMeta):
     # Referring to the pandas nullable dtypes; not just woodwork logical types
     _integer_nullable_incompatibilities = []
     _boolean_nullable_incompatibilities = []
-    is_multiseries = False
 
     def __init__(self, parameters=None, component_obj=None, random_seed=0, **kwargs):
         """Base class for all components.

evalml/pipelines/components/estimators/regressors/multiseries_time_series_baseline_regressor.py

@@ -25,12 +25,11 @@ class MultiseriesTimeSeriesBaselineRegressor(Estimator):
     """{}"""
     model_family = ModelFamily.BASELINE
     """ModelFamily.BASELINE"""
-    is_multiseries = True
     supported_problem_types = [
-        ProblemTypes.TIME_SERIES_REGRESSION,
+        ProblemTypes.MULTISERIES_TIME_SERIES_REGRESSION,
     ]
     """[
-        ProblemTypes.TIME_SERIES_REGRESSION,
+        ProblemTypes.MULTISERIES_TIME_SERIES_REGRESSION,
     ]"""
 
     def __init__(self, gap=1, forecast_horizon=1, random_seed=0, **kwargs):

evalml/pipelines/components/estimators/regressors/varmax_regressor.py

@@ -48,10 +48,9 @@ class VARMAXRegressor(Estimator):
         "trend": Categorical(['n', 'c', 't', 'ct']),
     }"""
     model_family = ModelFamily.VARMAX
-    is_multiseries = True
     """ModelFamily.VARMAX"""
-    supported_problem_types = [ProblemTypes.TIME_SERIES_REGRESSION]
-    """[ProblemTypes.TIME_SERIES_REGRESSION]"""
+    supported_problem_types = [ProblemTypes.MULTISERIES_TIME_SERIES_REGRESSION]
+    """[ProblemTypes.MULTISERIES_TIME_SERIES_REGRESSION]"""
 
     def __init__(
         self,

evalml/pipelines/components/utils.py

@@ -14,7 +14,7 @@
 from evalml.pipelines.components.component_base import ComponentBase
 from evalml.pipelines.components.estimators.estimator import Estimator
 from evalml.pipelines.components.transformers.transformer import Transformer
-from evalml.problem_types import ProblemTypes, handle_problem_types, is_time_series
+from evalml.problem_types import ProblemTypes, handle_problem_types
 from evalml.utils import get_importable_subclasses
 
 
@@ -56,18 +56,10 @@ def allowed_model_families(problem_type):
     return list(set([e.model_family for e in estimators]))
 
 
-def _filter_multiseries_estimators(estimators, is_multiseries):
-    if is_multiseries:
-        return [estimator for estimator in estimators if estimator.is_multiseries]
-    else:
-        return [estimator for estimator in estimators if not estimator.is_multiseries]
-
-
 def get_estimators(
     problem_type,
     model_families=None,
     excluded_model_families=None,
-    is_multiseries=False,
 ):
     """Returns the estimators allowed for a particular problem type.
 
@@ -77,7 +69,6 @@ def get_estimators(
         problem_type (ProblemTypes or str): Problem type to filter for.
         model_families (list[ModelFamily] or list[str]): Model families to filter for.
         excluded_model_families (list[ModelFamily]): A list of model families to exclude from the results.
-        is_multiseries (bool): Whether to return only estimators that support multiseries data.
 
     Returns:
         list[class]: A list of estimator subclasses.
@@ -124,11 +115,6 @@ def get_estimators(
         if estimator_class.model_family not in model_families:
             continue
         estimator_classes.append(estimator_class)
-    if is_time_series(problem_type):
-        estimator_classes = _filter_multiseries_estimators(
-            estimator_classes,
-            is_multiseries,
-        )
     return estimator_classes
 
 

evalml/pipelines/multiseries_regression_pipeline.py

@@ -21,9 +21,9 @@ class MultiseriesRegressionPipeline(TimeSeriesRegressionPipeline):
 
     """
 
-    problem_type = ProblemTypes.TIME_SERIES_REGRESSION
+    problem_type = ProblemTypes.MULTISERIES_TIME_SERIES_REGRESSION
 
-    """ProblemTypes.TIME_SERIES_REGRESSION"""
+    """ProblemTypes.MULTISERIES_TIME_SERIES_REGRESSION"""
 
     def __init__(
         self,