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.

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

Add new dataflow plan nodes for custom offset windows #1579

Open
wants to merge 13 commits into
base: main
Choose a base branch
from
Open
Show file tree
Hide file tree
Changes from all commits
Commits
File filter

Filter by extension

Filter by extension

Conversations
Failed to load comments.
Loading
Jump to
The table of contents is too big for display.
Diff view
Diff view
  •  
  •  
  •  
5 changes: 5 additions & 0 deletions metricflow-semantics/metricflow_semantics/dag/id_prefix.py
Original file line number Diff line number Diff line change
Expand Up @@ -56,6 +56,8 @@ class StaticIdPrefix(IdPrefix, Enum, metaclass=EnumMetaClassHelper):
DATAFLOW_NODE_JOIN_CONVERSION_EVENTS_PREFIX = "jce"
DATAFLOW_NODE_WINDOW_REAGGREGATION_ID_PREFIX = "wr"
DATAFLOW_NODE_ALIAS_SPECS_ID_PREFIX = "as"
DATAFLOW_NODE_CUSTOM_GRANULARITY_BOUNDS_ID_PREFIX = "cgb"
DATAFLOW_NODE_OFFSET_BY_CUSTOMG_GRANULARITY_ID_PREFIX = "obcg"
Copy link
Contributor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Nit: Typo.


SQL_EXPR_COLUMN_REFERENCE_ID_PREFIX = "cr"
SQL_EXPR_COMPARISON_ID_PREFIX = "cmp"
Expand All @@ -75,6 +77,9 @@ class StaticIdPrefix(IdPrefix, Enum, metaclass=EnumMetaClassHelper):
SQL_EXPR_BETWEEN_PREFIX = "betw"
SQL_EXPR_WINDOW_FUNCTION_ID_PREFIX = "wfnc"
SQL_EXPR_GENERATE_UUID_PREFIX = "uuid"
SQL_EXPR_CASE_PREFIX = "case"
SQL_EXPR_ARITHMETIC_PREFIX = "arit"
SQL_EXPR_INTEGER_PREFIX = "int"

SQL_PLAN_SELECT_STATEMENT_ID_PREFIX = "ss"
SQL_PLAN_TABLE_FROM_CLAUSE_ID_PREFIX = "tfc"
Expand Down
6 changes: 1 addition & 5 deletions metricflow-semantics/metricflow_semantics/instances.py
Original file line number Diff line number Diff line change
Expand Up @@ -164,11 +164,7 @@ def with_entity_prefix(
) -> TimeDimensionInstance:
"""Returns a new instance with the entity prefix added to the entity links."""
transformed_spec = self.spec.with_entity_prefix(entity_prefix)
return TimeDimensionInstance(
associated_columns=(column_association_resolver.resolve_spec(transformed_spec),),
defined_from=self.defined_from,
spec=transformed_spec,
)
return self.with_new_spec(transformed_spec, column_association_resolver)

def with_new_defined_from(self, defined_from: Sequence[SemanticModelElementReference]) -> TimeDimensionInstance:
"""Returns a new instance with the defined_from field replaced."""
Expand Down
Original file line number Diff line number Diff line change
@@ -1,6 +1,5 @@
from __future__ import annotations

from metricflow_semantics.model.semantic_manifest_lookup import SemanticManifestLookup
from metricflow_semantics.naming.linkable_spec_name import DUNDER
from metricflow_semantics.specs.column_assoc import (
ColumnAssociation,
Expand Down Expand Up @@ -28,8 +27,8 @@ class DunderColumnAssociationResolver(ColumnAssociationResolver):
listing__country
"""

def __init__(self, semantic_manifest_lookup: SemanticManifestLookup) -> None: # noqa: D107
self._visitor_helper = DunderColumnAssociationResolverVisitor(semantic_manifest_lookup)
def __init__(self) -> None: # noqa: D107
self._visitor_helper = DunderColumnAssociationResolverVisitor()

def resolve_spec(self, spec: InstanceSpec) -> ColumnAssociation: # noqa: D102
return spec.accept(self._visitor_helper)
Expand All @@ -38,9 +37,6 @@ def resolve_spec(self, spec: InstanceSpec) -> ColumnAssociation: # noqa: D102
class DunderColumnAssociationResolverVisitor(InstanceSpecVisitor[ColumnAssociation]):
"""Visitor helper class for DefaultColumnAssociationResolver2."""

def __init__(self, semantic_manifest_lookup: SemanticManifestLookup) -> None: # noqa: D107
self._semantic_manifest_lookup = semantic_manifest_lookup

def visit_metric_spec(self, metric_spec: MetricSpec) -> ColumnAssociation: # noqa: D102
return ColumnAssociation(metric_spec.element_name if metric_spec.alias is None else metric_spec.alias)

Expand All @@ -58,6 +54,11 @@ def visit_time_dimension_spec(self, time_dimension_spec: TimeDimensionSpec) -> C
if time_dimension_spec.aggregation_state
else ""
)
+ (
f"{DUNDER}{time_dimension_spec.window_function.value.lower()}"
if time_dimension_spec.window_function
else ""
)
)

def visit_entity_spec(self, entity_spec: EntitySpec) -> ColumnAssociation: # noqa: D102
Expand Down
Original file line number Diff line number Diff line change
Expand Up @@ -15,6 +15,7 @@
from metricflow_semantics.naming.linkable_spec_name import StructuredLinkableSpecName
from metricflow_semantics.specs.dimension_spec import DimensionSpec
from metricflow_semantics.specs.instance_spec import InstanceSpecVisitor
from metricflow_semantics.sql.sql_exprs import SqlWindowFunction
from metricflow_semantics.time.granularity import ExpandedTimeGranularity
from metricflow_semantics.visitor import VisitorOutputT

Expand Down Expand Up @@ -91,6 +92,8 @@ class TimeDimensionSpec(DimensionSpec): # noqa: D101
# Used for semi-additive joins. Some more thought is needed, but this may be useful in InstanceSpec.
aggregation_state: Optional[AggregationState] = None

window_function: Optional[SqlWindowFunction] = None

@property
def without_first_entity_link(self) -> TimeDimensionSpec: # noqa: D102
assert len(self.entity_links) > 0, f"Spec does not have any entity links: {self}"
Expand All @@ -99,6 +102,8 @@ def without_first_entity_link(self) -> TimeDimensionSpec: # noqa: D102
entity_links=self.entity_links[1:],
time_granularity=self.time_granularity,
date_part=self.date_part,
aggregation_state=self.aggregation_state,
window_function=self.window_function,
)

@property
Expand All @@ -108,6 +113,8 @@ def without_entity_links(self) -> TimeDimensionSpec: # noqa: D102
time_granularity=self.time_granularity,
date_part=self.date_part,
entity_links=(),
aggregation_state=self.aggregation_state,
window_function=self.window_function,
)

@property
Expand Down Expand Up @@ -153,6 +160,7 @@ def with_grain(self, time_granularity: ExpandedTimeGranularity) -> TimeDimension
time_granularity=time_granularity,
date_part=self.date_part,
aggregation_state=self.aggregation_state,
window_function=self.window_function,
)

def with_base_grain(self) -> TimeDimensionSpec: # noqa: D102
Expand All @@ -162,6 +170,7 @@ def with_base_grain(self) -> TimeDimensionSpec: # noqa: D102
time_granularity=ExpandedTimeGranularity.from_time_granularity(self.time_granularity.base_granularity),
date_part=self.date_part,
aggregation_state=self.aggregation_state,
window_function=self.window_function,
)

def with_grain_and_date_part( # noqa: D102
Expand All @@ -173,6 +182,7 @@ def with_grain_and_date_part( # noqa: D102
time_granularity=time_granularity,
date_part=date_part,
aggregation_state=self.aggregation_state,
window_function=self.window_function,
)

def with_aggregation_state(self, aggregation_state: AggregationState) -> TimeDimensionSpec: # noqa: D102
Expand All @@ -182,6 +192,17 @@ def with_aggregation_state(self, aggregation_state: AggregationState) -> TimeDim
time_granularity=self.time_granularity,
date_part=self.date_part,
aggregation_state=aggregation_state,
window_function=self.window_function,
)

def with_window_function(self, window_function: SqlWindowFunction) -> TimeDimensionSpec: # noqa: D102
return TimeDimensionSpec(
element_name=self.element_name,
entity_links=self.entity_links,
time_granularity=self.time_granularity,
date_part=self.date_part,
aggregation_state=self.aggregation_state,
window_function=window_function,
)

def comparison_key(self, exclude_fields: Sequence[TimeDimensionSpecField] = ()) -> TimeDimensionSpecComparisonKey:
Expand Down Expand Up @@ -243,6 +264,7 @@ def with_entity_prefix(self, entity_prefix: EntityReference) -> TimeDimensionSpe
time_granularity=self.time_granularity,
date_part=self.date_part,
aggregation_state=self.aggregation_state,
window_function=self.window_function,
)

@staticmethod
Expand Down
Loading
Loading