import copy
import itertools
import logging
import os
import threading
import typing
import warnings
from collections import Counter, defaultdict
from datetime import datetime, timedelta, timezone
from pathlib import Path
from typing import (
Any,
Dict,
Iterable,
List,
Mapping,
Optional,
Sequence,
Set,
Tuple,
Union,
cast,
)
import pandas as pd
import pyarrow
from dateutil.tz import tzlocal
from google.protobuf.timestamp_pb2 import Timestamp
from feast.aggregation import aggregation_specs_to_agg_ops
from feast.constants import FEAST_FS_YAML_FILE_PATH_ENV_NAME
from feast.entity import Entity
from feast.errors import (
FeatureNameCollisionError,
FeatureViewNotFoundException,
RequestDataNotFoundInEntityRowsException,
)
from feast.field import Field
from feast.infra.compute_engines.backends.pandas_backend import PandasBackend
from feast.infra.key_encoding_utils import deserialize_entity_key
from feast.protos.feast.serving.ServingService_pb2 import (
FieldStatus,
GetOnlineFeaturesResponse,
)
from feast.protos.feast.types.EntityKey_pb2 import EntityKey as EntityKeyProto
from feast.protos.feast.types.Value_pb2 import FloatList as FloatListProto
from feast.protos.feast.types.Value_pb2 import RepeatedValue as RepeatedValueProto
from feast.protos.feast.types.Value_pb2 import Value as ValueProto
from feast.type_map import python_values_to_proto_values
from feast.types import ComplexFeastType, PrimitiveFeastType
from feast.value_type import ValueType
from feast.version import get_version
if typing.TYPE_CHECKING:
from feast.base_feature_view import BaseFeatureView
from feast.feature_service import FeatureService
from feast.feature_view import FeatureView
from feast.infra.registry.base_registry import BaseRegistry
from feast.on_demand_feature_view import OnDemandFeatureView
APPLICATION_NAME = "feast-dev/feast"
USER_AGENT = "{}/{}".format(APPLICATION_NAME, get_version())
def _parse_feature_ref(ref: str) -> Tuple[str, Optional[int], str]:
"""Parse 'fv_name@version:feature' into (fv_name, version_number, feature_name).
If no @version is present, version_number is None (meaning 'latest').
Examples:
'driver_stats:trips' -> ('driver_stats', None, 'trips')
'driver_stats@v2:trips' -> ('driver_stats', 2, 'trips')
'driver_stats@latest:trips' -> ('driver_stats', None, 'trips')
"""
import re
colon_idx = ref.find(":")
if colon_idx < 0:
raise ValueError(
f"Invalid feature reference '{ref}'. Expected format: ':' "
f"or '@:'"
)
fv_part = ref[:colon_idx]
feature_name = ref[colon_idx + 1 :]
at_idx = fv_part.find("@")
if at_idx < 0:
return (fv_part, None, feature_name)
fv_name = fv_part[:at_idx]
version_str = fv_part[at_idx + 1 :]
if not version_str or version_str.lower() == "latest":
return (fv_name, None, feature_name)
# Parse version number from formats like "v2", "V2"
match = re.match(r"^[vV](\d+)$", version_str)
if not match:
# Not a recognized version format treat entire fv_part as the name
return (fv_part, None, feature_name)
return (fv_name, int(match.group(1)), feature_name)
def _strip_version_from_ref(ref: str) -> str:
"""Strip @version from a feature reference, returning 'fv_name:feature'.
Used to produce clean refs for output column naming.
"""
fv_name, _, feature_name = _parse_feature_ref(ref)
return f"{fv_name}:{feature_name}"
def get_user_agent():
return USER_AGENT
def make_tzaware(t: datetime) -> datetime:
"""We assume tz-naive datetimes are UTC"""
if t.tzinfo is None:
return t.replace(tzinfo=timezone.utc)
else:
return t
def compute_non_entity_date_range(
feature_views: List["FeatureView"],
start_date: Optional[datetime] = None,
end_date: Optional[datetime] = None,
default_window_days: int = 30,
) -> Tuple[datetime, datetime]:
if end_date is None:
end_date = datetime.now(tz=timezone.utc)
else:
end_date = make_tzaware(end_date)
if start_date is None:
max_ttl_seconds = max(
(
int(fv.ttl.total_seconds())
for fv in feature_views
if fv.ttl and isinstance(fv.ttl, timedelta)
),
default=0,
)
start_date = end_date - timedelta(
seconds=max_ttl_seconds
if max_ttl_seconds > 0
else default_window_days * 86400
)
else:
start_date = make_tzaware(start_date)
return start_date, end_date
def make_df_tzaware(t: pd.DataFrame) -> pd.DataFrame:
"""Make all datetime type columns tzaware; leave everything else intact."""
df = t.copy() # don't modify incoming dataframe inplace
for column in df.columns:
if pd.api.types.is_datetime64_any_dtype(df[column]):
df[column] = pd.to_datetime(df[column], utc=True)
return df
def to_naive_utc(ts: datetime) -> datetime:
if ts.tzinfo is None:
return ts
else:
return ts.astimezone(timezone.utc).replace(tzinfo=None)
def maybe_local_tz(t: datetime) -> datetime:
if t.tzinfo is None:
return t.replace(tzinfo=tzlocal())
else:
return t
def get_default_yaml_file_path(repo_path: Path) -> Path:
if FEAST_FS_YAML_FILE_PATH_ENV_NAME in os.environ:
yaml_path = os.environ[FEAST_FS_YAML_FILE_PATH_ENV_NAME]
return Path(yaml_path)
else:
return repo_path / "feature_store.yaml"
def _get_requested_feature_views_to_features_dict(
feature_refs: List[str],
feature_views: List["FeatureView"],
on_demand_feature_views: List["OnDemandFeatureView"],
) -> Tuple[Dict["FeatureView", List[str]], Dict["OnDemandFeatureView", List[str]]]:
"""Create a dict of FeatureView -> List[Feature] for all requested features.
Set full_feature_names to True to have feature names prefixed by their feature view name.
"""
feature_views_to_feature_map: Dict["FeatureView", List[str]] = defaultdict(list)
on_demand_feature_views_to_feature_map: Dict["OnDemandFeatureView", List[str]] = (
defaultdict(list)
)
for ref in feature_refs:
fv_name, version_num, feature_from_ref = _parse_feature_ref(ref)
# Build the key that matches projection.name_to_use()
if version_num is not None:
feature_view_from_ref = f"{fv_name}@v{version_num}"
else:
feature_view_from_ref = fv_name
found = False
for fv in feature_views:
if fv.projection.name_to_use() == feature_view_from_ref:
found = True
feature_views_to_feature_map[fv].append(feature_from_ref)
for odfv in on_demand_feature_views:
if odfv.projection.name_to_use() == feature_view_from_ref:
found = True
on_demand_feature_views_to_feature_map[odfv].append(feature_from_ref)
if not found:
raise ValueError(f"Could not find feature view from reference {ref}")
return feature_views_to_feature_map, on_demand_feature_views_to_feature_map
def _get_column_names(
feature_view: "FeatureView", entities: List[Entity]
) -> Tuple[List[str], List[str], str, Optional[str]]:
"""
If a field mapping exists, run it in reverse on the join keys,
feature names, event timestamp column, and created timestamp column
to get the names of the relevant columns in the offline feature store table.
Returns:
Tuple containing the list of reverse-mapped join_keys,
reverse-mapped feature names, reverse-mapped event timestamp column,
and reverse-mapped created timestamp column that will be passed into
the query to the offline store.
"""
if feature_view.batch_source is None:
raise ValueError(
f"Feature view '{feature_view.name}' has no batch_source and cannot be used for offline retrieval."
)
# if we have mapped fields, use the original field names in the call to the offline store
timestamp_field = feature_view.batch_source.timestamp_field
# For feature views with aggregations, read INPUT columns from aggregations.
# This applies to StreamFeatureView, BatchFeatureView,
# or any FeatureView that has aggregations.
if hasattr(feature_view, "aggregations") and feature_view.aggregations:
# Extract unique input columns from aggregations, preserving order
feature_names = list(
dict.fromkeys(agg.column for agg in feature_view.aggregations)
)
else:
# For regular feature views, use the feature names
feature_names = [feature.name for feature in feature_view.features]
created_timestamp_column = feature_view.batch_source.created_timestamp_column
from feast.feature_view import DUMMY_ENTITY_ID
join_keys = [
entity.join_key for entity in entities if entity.join_key != DUMMY_ENTITY_ID
]
if feature_view.batch_source.field_mapping is not None:
reverse_field_mapping = {
v: k for k, v in feature_view.batch_source.field_mapping.items()
}
timestamp_field = (
reverse_field_mapping[timestamp_field]
if timestamp_field in reverse_field_mapping.keys()
else timestamp_field
)
created_timestamp_column = (
reverse_field_mapping[created_timestamp_column]
if created_timestamp_column
and created_timestamp_column in reverse_field_mapping.keys()
else created_timestamp_column
)
join_keys = [
reverse_field_mapping[col] if col in reverse_field_mapping.keys() else col
for col in join_keys
]
feature_names = [
reverse_field_mapping[col] if col in reverse_field_mapping.keys() else col
for col in feature_names
]
# We need to exclude join keys and timestamp columns from the list of features, after they are mapped to
# their final column names via the `field_mapping` field of the source.
feature_names = [
name
for name in feature_names
if name not in join_keys
and name != timestamp_field
and name != created_timestamp_column
]
return (
join_keys,
feature_names,
timestamp_field,
created_timestamp_column,
)
def _run_pyarrow_field_mapping(
table: pyarrow.Table,
field_mapping: Dict[str, str],
) -> pyarrow.Table:
# run field mapping in the forward direction
cols = table.column_names
mapped_cols = [
field_mapping[col] if col in field_mapping.keys() else col for col in cols
]
table = table.rename_columns(mapped_cols)
return table
def _get_fields_with_aliases(
fields: List[str],
field_mappings: Dict[str, str],
) -> Tuple[List[str], List[str]]:
"""
Get a list of fields with aliases based on the field mappings.
"""
for field in fields:
if "." in field and field not in field_mappings:
raise ValueError(
f"Feature {field} contains a '.' character, which is not allowed in field names. Use field mappings to rename fields."
)
fields_with_aliases = [
f"{field} AS {field_mappings[field]}" if field in field_mappings else field
for field in fields
]
aliases = [
field_mappings[field] if field in field_mappings else field for field in fields
]
return (fields_with_aliases, aliases)
def _coerce_datetime(ts):
"""
Depending on underlying time resolution, arrow to_pydict() sometimes returns pd
timestamp type (for nanosecond resolution), and sometimes you get standard python datetime
(for microsecond resolution).
While pd timestamp class is a subclass of python datetime, it doesn't always behave the
same way. We convert it to normal datetime so that consumers downstream don't have to deal
with these quirks.
"""
if isinstance(ts, pd.Timestamp):
return ts.to_pydatetime()
else:
return ts
def _columns_to_proto_values(
table: pyarrow.RecordBatch,
columns: List[Tuple[str, ValueType]],
allow_missing: bool = False,
) -> Dict[str, List[ValueProto]]:
"""Convert table columns to proto values dict.
Args:
table: PyArrow RecordBatch containing the data.
columns: List of (column_name, value_type) tuples to convert.
allow_missing: If True, skip columns not found in table. If False, raise ValueError.
Returns:
Dict mapping column names to lists of ValueProto.
"""
result: Dict[str, List[ValueProto]] = {}
for column, value_type in columns:
if column in table.column_names:
result[column] = python_values_to_proto_values(
table.column(column).to_numpy(zero_copy_only=False), value_type
)
elif not allow_missing:
raise ValueError(f"Column {column} not found in table")
return result
def _build_entity_keys(
num_rows: int,
join_keys: Dict[str, ValueType],
proto_values: Dict[str, List[ValueProto]],
) -> List[EntityKeyProto]:
"""Build entity key protos for each row.
Args:
num_rows: Number of rows to generate entity keys for.
join_keys: Dict mapping join key names to their value types.
proto_values: Dict mapping column names to lists of ValueProto values.
Returns:
List of EntityKeyProto, one per row.
"""
return [
EntityKeyProto(
join_keys=list(join_keys.keys()),
entity_values=[
proto_values[k][idx] for k in join_keys if k in proto_values
],
)
for idx in range(num_rows)
]
def _convert_arrow_to_proto(
table: Union[pyarrow.Table, pyarrow.RecordBatch],
feature_view: Union["FeatureView", "BaseFeatureView", "OnDemandFeatureView"],
join_keys: Dict[str, ValueType],
) -> List[Tuple[EntityKeyProto, Dict[str, ValueProto], datetime, Optional[datetime]]]:
# This is a workaround for isinstance(feature_view, OnDemandFeatureView), which triggers a circular import
# Check for source_request_sources or source_feature_view_projections attributes to identify ODFVs
if (
getattr(feature_view, "source_request_sources", None) is not None
or getattr(feature_view, "source_feature_view_projections", None) is not None
):
return _convert_arrow_odfv_to_proto(table, feature_view, join_keys) # type: ignore[arg-type]
else:
return _convert_arrow_fv_to_proto(table, feature_view, join_keys) # type: ignore[arg-type]
def _convert_arrow_fv_to_proto(
table: Union[pyarrow.Table, pyarrow.RecordBatch],
feature_view: "FeatureView",
join_keys: Dict[str, ValueType],
) -> List[Tuple[EntityKeyProto, Dict[str, ValueProto], datetime, Optional[datetime]]]:
# Avoid ChunkedArrays which guarantees `zero_copy_only` available.
if isinstance(table, pyarrow.Table):
table = table.to_batches()[0]
if feature_view.batch_source is None:
raise ValueError(
f"Feature view '{feature_view.name}' has no batch_source and cannot be converted to proto."
)
# TODO: This will break if the feature view has aggregations or transformations
columns = [
(field.name, field.dtype.to_value_type()) for field in feature_view.features
] + list(join_keys.items())
proto_values_by_column = _columns_to_proto_values(
table, columns, allow_missing=False
)
entity_keys = _build_entity_keys(table.num_rows, join_keys, proto_values_by_column)
# Serialize the features per row
feature_dict = {
feature.name: proto_values_by_column[feature.name]
for feature in feature_view.features
}
features = [dict(zip(feature_dict, vars)) for vars in zip(*feature_dict.values())]
# Convert event_timestamps
event_timestamps = [
_coerce_datetime(val)
for val in pd.to_datetime(
table.column(feature_view.batch_source.timestamp_field).to_numpy(
zero_copy_only=False
)
)
]
# Convert created_timestamps if they exist
if feature_view.batch_source.created_timestamp_column:
created_timestamps = [
_coerce_datetime(val)
for val in pd.to_datetime(
table.column(
feature_view.batch_source.created_timestamp_column
).to_numpy(zero_copy_only=False)
)
]
else:
created_timestamps = [None] * table.num_rows
return list(zip(entity_keys, features, event_timestamps, created_timestamps))
def _convert_arrow_odfv_to_proto(
table: Union[pyarrow.Table, pyarrow.RecordBatch],
feature_view: "OnDemandFeatureView",
join_keys: Dict[str, ValueType],
) -> List[Tuple[EntityKeyProto, Dict[str, ValueProto], datetime, Optional[datetime]]]:
# Avoid ChunkedArrays which guarantees `zero_copy_only` available.
if isinstance(table, pyarrow.Table):
table = table.to_batches()[0]
columns = [
(field.name, field.dtype.to_value_type()) for field in feature_view.features
] + list(join_keys.items())
# Convert columns that exist in the table
proto_values_by_column = _columns_to_proto_values(
table, columns, allow_missing=True
)
# Ensure join keys are included, creating null values if missing from table
for join_key, value_type in join_keys.items():
if join_key not in proto_values_by_column:
if join_key in table.column_names:
proto_values_by_column[join_key] = python_values_to_proto_values(
table.column(join_key).to_numpy(zero_copy_only=False), value_type
)
else:
# Create null proto values directly (no need to build a PyArrow array)
proto_values_by_column[join_key] = python_values_to_proto_values(
[None] * table.num_rows, value_type
)
# Cache column names set to avoid recreating list in loop
column_names = {c[0] for c in columns}
# Adding On Demand Features that are missing from proto_values
for feature in feature_view.features:
if feature.name in column_names and feature.name not in proto_values_by_column:
# Create null proto values directly (more efficient than building PyArrow array)
proto_values_by_column[feature.name] = python_values_to_proto_values(
[None] * table.num_rows, feature.dtype.to_value_type()
)
entity_keys = _build_entity_keys(table.num_rows, join_keys, proto_values_by_column)
# Serialize the features per row
feature_dict = {
feature.name: proto_values_by_column[feature.name]
for feature in feature_view.features
if feature.name in proto_values_by_column
}
if feature_view.write_to_online_store:
table_columns = {col.name for col in table.schema}
for feature in feature_view.schema:
if feature.name not in feature_dict and feature.name in table_columns:
feature_dict[feature.name] = proto_values_by_column[feature.name]
features = [dict(zip(feature_dict, vars)) for vars in zip(*feature_dict.values())]
# We need to artificially add event_timestamps and created_timestamps
now = _utc_now()
event_timestamps = [
_coerce_datetime(pd.Timestamp(now)) for _ in range(table.num_rows)
]
# setting them equivalent
created_timestamps = event_timestamps
return list(zip(entity_keys, features, event_timestamps, created_timestamps))
def _validate_entity_values(join_key_values: Dict[str, List[ValueProto]]):
set_of_row_lengths = {len(v) for v in join_key_values.values()}
if len(set_of_row_lengths) > 1:
raise ValueError("All entity rows must have the same columns.")
return set_of_row_lengths.pop()
def _validate_feature_refs(feature_refs: List[str], full_feature_names: bool = False):
"""
Validates that there are no collisions among the feature references.
Args:
feature_refs: List of feature references to validate. Feature references must have format
"feature_view:feature", e.g. "customer_fv:daily_transactions".
full_feature_names: If True, the full feature references are compared for collisions; if False,
only the feature names are compared.
Raises:
FeatureNameCollisionError: There is a collision among the feature references.
"""
collided_feature_refs = []
if full_feature_names:
collided_feature_refs = [
ref for ref, occurrences in Counter(feature_refs).items() if occurrences > 1
]
else:
feature_names = [_parse_feature_ref(ref)[2] for ref in feature_refs]
collided_feature_names = [
ref
for ref, occurrences in Counter(feature_names).items()
if occurrences > 1
]
for feature_name in collided_feature_names:
collided_feature_refs.extend(
[ref for ref in feature_refs if ref.endswith(":" + feature_name)]
)
if len(collided_feature_refs) > 0:
raise FeatureNameCollisionError(collided_feature_refs, full_feature_names)
def _group_feature_refs(
features: List[str],
all_feature_views: List["FeatureView"],
all_on_demand_feature_views: List["OnDemandFeatureView"],
) -> Tuple[
List[Tuple[Union["FeatureView", "OnDemandFeatureView"], List[str]]],
List[Tuple["OnDemandFeatureView", List[str]]],
]:
"""Get list of feature views and corresponding feature names based on feature references"""
# view name to view proto
view_index: Dict[str, Union["FeatureView", "OnDemandFeatureView"]] = {
view.projection.name_to_use(): view for view in all_feature_views
}
# on demand view to on demand view proto
on_demand_view_index: Dict[str, "OnDemandFeatureView"] = {}
for view in all_on_demand_feature_views:
if view.projection and not view.write_to_online_store:
on_demand_view_index[view.projection.name_to_use()] = view
elif view.projection and view.write_to_online_store:
# we insert the ODFV view to FVs for ones that are written to the online store
view_index[view.projection.name_to_use()] = view
# view name to feature names
views_features = defaultdict(set)
# on demand view name to feature names
on_demand_view_features = defaultdict(set)
for ref in features:
fv_name, version_num, feat_name = _parse_feature_ref(ref)
# Build the key that matches projection.name_to_use()
if version_num is not None:
view_name = f"{fv_name}@v{version_num}"
else:
view_name = fv_name
if view_name in view_index:
if hasattr(view_index[view_name], "write_to_online_store"):
tmp_feat_name = [
f for f in view_index[view_name].schema if f.name == feat_name
]
if len(tmp_feat_name) > 0:
feat_name = tmp_feat_name[0].name
else:
view_index[view_name].projection.get_feature(
feat_name
) # For validation
views_features[view_name].add(feat_name)
elif view_name in on_demand_view_index:
on_demand_view_index[view_name].projection.get_feature(
feat_name
) # For validation
on_demand_view_features[view_name].add(feat_name)
# Let's also add in any FV Feature dependencies here.
for input_fv_projection in on_demand_view_index[
view_name
].source_feature_view_projections.values():
for input_feat in input_fv_projection.features:
views_features[input_fv_projection.name].add(input_feat.name)
else:
raise FeatureViewNotFoundException(view_name)
fvs_result: List[Tuple[Union["FeatureView", "OnDemandFeatureView"], List[str]]] = []
odfvs_result: List[Tuple["OnDemandFeatureView", List[str]]] = []
for view_name, feature_names in views_features.items():
fvs_result.append((view_index[view_name], list(feature_names)))
for view_name, feature_names in on_demand_view_features.items():
odfvs_result.append((on_demand_view_index[view_name], list(feature_names)))
return fvs_result, odfvs_result
def _apply_aggregations_to_response(
response_data: Union[pyarrow.Table, Dict[str, List[Any]]],
aggregations,
group_keys: Optional[List[str]],
mode: str,
) -> Union[pyarrow.Table, Dict[str, List[Any]]]:
"""
Apply aggregations using PandasBackend.
Args:
response_data: Either a pyarrow.Table or dict of lists containing the data
aggregations: List of Aggregation objects to apply
group_keys: List of column names to group by (optional)
mode: Transformation mode ("python", "pandas", or "substrait")
Returns:
Aggregated data in the same format as input
TODO: Consider refactoring to support backends other than pandas in the future.
"""
if not aggregations:
return response_data
backend = PandasBackend()
# Convert to pandas DataFrame
if isinstance(response_data, dict):
df = pd.DataFrame(response_data)
else: # pyarrow.Table
df = backend.from_arrow(response_data)
if df.empty:
return response_data
# Convert aggregations to agg_ops format
agg_ops = aggregation_specs_to_agg_ops(
aggregations,
time_window_unsupported_error_message=(
"Time window aggregation is not supported in online serving."
),
)
# Apply aggregations using PandasBackend
if group_keys:
result_df = backend.groupby_agg(df, group_keys, agg_ops)
else:
# No grouping - aggregate over entire dataset
result_df = backend.groupby_agg(df, [], agg_ops)
# Convert back to original format
if mode == "python":
return {col: result_df[col].tolist() for col in result_df.columns}
else: # pandas or substrait
return backend.to_arrow(result_df)
def _augment_response_with_on_demand_transforms(
online_features_response: GetOnlineFeaturesResponse,
feature_refs: List[str],
requested_on_demand_feature_views: List["OnDemandFeatureView"],
full_feature_names: bool,
feature_types: Optional[Dict[str, "ValueType"]] = None,
):
"""Computes on demand feature values and adds them to the result rows.
Assumes that 'online_features_response' already contains the necessary request data and input feature
views for the on demand feature views. Unneeded feature values such as request data and
unrequested input feature views will be removed from 'online_features_response'.
Args:
online_features_response: Protobuf object to populate
feature_refs: List of all feature references to be returned.
requested_on_demand_feature_views: List of all odfvs that have been requested.
full_feature_names: A boolean that provides the option to add the feature view prefixes to the feature names,
changing them from the format "feature" to "feature_view__feature" (e.g., "daily_transactions" changes to
"customer_fv__daily_transactions").
"""
from feast.online_response import OnlineResponse
requested_odfv_map = {odfv.name: odfv for odfv in requested_on_demand_feature_views}
requested_odfv_feature_names = requested_odfv_map.keys()
odfv_feature_refs = defaultdict(list)
for feature_ref in feature_refs:
view_name, _, feature_name = _parse_feature_ref(feature_ref)
if view_name in requested_odfv_feature_names:
odfv_feature_refs[view_name].append(
f"{requested_odfv_map[view_name].projection.name_to_use()}__{feature_name}"
if full_feature_names
else feature_name
)
initial_response = OnlineResponse(
online_features_response, feature_types=feature_types
)
initial_response_arrow: Optional[pyarrow.Table] = None
initial_response_dict: Optional[Dict[str, List[Any]]] = None
def _is_metrics_active():
try:
from feast.metrics import _config
return _config.online_features
except Exception:
return False
_metrics_active = _is_metrics_active()
# Apply on demand transformations and augment the result rows
odfv_result_names = set()
for odfv_name, _feature_refs in odfv_feature_refs.items():
odfv = requested_odfv_map[odfv_name]
if not odfv.write_to_online_store:
_should_track = _metrics_active and getattr(odfv, "track_metrics", False)
if _should_track:
import time as _time
_transform_start = _time.monotonic()
# Apply aggregations if configured.
if odfv.aggregations:
if odfv.mode == "python":
if initial_response_dict is None:
initial_response_dict = initial_response.to_dict()
initial_response_dict = _apply_aggregations_to_response(
initial_response_dict,
odfv.aggregations,
odfv.entities,
odfv.mode,
)
elif odfv.mode in {"pandas", "substrait"}:
if initial_response_arrow is None:
initial_response_arrow = initial_response.to_arrow()
initial_response_arrow = _apply_aggregations_to_response(
initial_response_arrow,
odfv.aggregations,
odfv.entities,
odfv.mode,
)
continue
# Apply transformation. Note: aggregations and transformation configs are mutually exclusive
# TODO: Fix to make it work for having both aggregation and transformation
# ticket: https://github.com/feast-dev/feast/issues/5689
if odfv.mode == "python":
if initial_response_dict is None:
initial_response_dict = initial_response.to_dict()
transformed_features_dict: Dict[str, List[Any]] = odfv.transform_dict(
initial_response_dict
)
elif odfv.mode in {"pandas", "substrait"}:
if initial_response_arrow is None:
initial_response_arrow = initial_response.to_arrow()
transformed_features_arrow = odfv.transform_arrow(
initial_response_arrow, full_feature_names
)
else:
raise Exception(
f"Invalid OnDemandFeatureMode: {odfv.mode}. Expected one of 'pandas', 'python', or 'substrait'."
)
if _should_track:
from feast.metrics import track_transformation
track_transformation(
odfv_name, odfv.mode, _time.monotonic() - _transform_start
)
transformed_features = (
transformed_features_dict
if odfv.mode == "python"
else transformed_features_arrow
)
transformed_columns = (
transformed_features.column_names
if isinstance(transformed_features, pyarrow.Table)
else transformed_features
)
selected_subset = [f for f in transformed_columns if f in _feature_refs]
proto_values = []
schema_dict = {k.name: k.dtype for k in odfv.schema}
for selected_feature in selected_subset:
feature_vector = transformed_features[selected_feature]
selected_feature_type = schema_dict.get(selected_feature, None)
feature_type: ValueType = ValueType.UNKNOWN
if selected_feature_type is not None:
if isinstance(
selected_feature_type, (ComplexFeastType, PrimitiveFeastType)
):
feature_type = selected_feature_type.to_value_type()
elif not isinstance(selected_feature_type, ValueType):
raise TypeError(
f"Unexpected type for feature_type: {type(feature_type)}"
)
proto_values.append(
python_values_to_proto_values(
feature_vector
if isinstance(feature_vector, list)
else [feature_vector]
if odfv.mode == "python"
else feature_vector.to_numpy(),
feature_type,
)
)
odfv_result_names |= set(selected_subset)
online_features_response.metadata.feature_names.val.extend(selected_subset)
for feature_idx in range(len(selected_subset)):
online_features_response.results.append(
GetOnlineFeaturesResponse.FeatureVector(
values=proto_values[feature_idx],
statuses=[FieldStatus.PRESENT] * len(proto_values[feature_idx]),
event_timestamps=[Timestamp()] * len(proto_values[feature_idx]),
)
)
def _get_entity_maps(
registry,
project,
feature_views,
) -> Tuple[Dict[str, str], Dict[str, ValueType], Set[str]]:
# TODO(felixwang9817): Support entities that have different types for different feature views.
entities = registry.list_entities(project, allow_cache=True)
entity_by_name: Dict[str, "Entity"] = {entity.name: entity for entity in entities}
entity_name_to_join_key_map: Dict[str, str] = {}
entity_type_map: Dict[str, ValueType] = {}
for entity in entities:
entity_name_to_join_key_map[entity.name] = entity.join_key
for feature_view in feature_views:
for entity_name in feature_view.entities:
entity = entity_by_name.get(entity_name)
if entity is None:
from feast.errors import EntityNotFoundException
raise EntityNotFoundException(entity_name, project=project)
# User directly uses join_key as the entity reference in the entity_rows for the
# entity mapping case.
entity_name = feature_view.projection.join_key_map.get(
entity.join_key, entity.name
)
join_key = feature_view.projection.join_key_map.get(
entity.join_key, entity.join_key
)
entity_name_to_join_key_map[entity_name] = join_key
for entity_column in feature_view.entity_columns:
dtype = entity_column.dtype.to_value_type()
entity_join_key_column_name = feature_view.projection.join_key_map.get(
entity_column.name, entity_column.name
)
entity_type_map[entity_join_key_column_name] = dtype
return (
entity_name_to_join_key_map,
entity_type_map,
set(entity_name_to_join_key_map.values()),
)
def _get_table_entity_values(
table: "FeatureView",
entity_name_to_join_key_map: Dict[str, str],
join_key_proto_values: Dict[str, List[ValueProto]],
) -> Dict[str, List[ValueProto]]:
# The correct join_keys expected by the OnlineStore for this Feature View.
table_join_keys = [
entity_name_to_join_key_map[entity_name] for entity_name in table.entities
]
# If the FeatureView has a Projection then the join keys may be aliased.
alias_to_join_key_map = {v: k for k, v in table.projection.join_key_map.items()}
# Subset to columns which are relevant to this FeatureView and
# give them the correct names.
entity_values = {
alias_to_join_key_map.get(k, k): v
for k, v in join_key_proto_values.items()
if alias_to_join_key_map.get(k, k) in table_join_keys
}
return entity_values
def _get_unique_entities(
table: "FeatureView",
join_key_values: Dict[str, List[ValueProto]],
entity_name_to_join_key_map: Dict[str, str],
) -> Tuple[Tuple[Dict[str, ValueProto], ...], Tuple[List[int], ...], int]:
"""Return the set of unique composite Entities for a Feature View and the indexes at which they appear.
This method allows us to query the OnlineStore for data we need only once
rather than requesting and processing data for the same combination of
Entities multiple times.
"""
# Get the correct set of entity values with the correct join keys.
table_entity_values = _get_table_entity_values(
table,
entity_name_to_join_key_map,
join_key_values,
)
# Validate that all expected join keys exist and have non-empty values.
expected_keys = set(entity_name_to_join_key_map.values())
expected_keys.discard("__dummy_id")
missing_keys = sorted(
list(set([key for key in expected_keys if key not in table_entity_values]))
)
empty_keys = sorted(
list(set([key for key in expected_keys if not table_entity_values.get(key)]))
)
if missing_keys or empty_keys:
if not any(table_entity_values.values()):
raise KeyError(
f"Missing join key values for keys: {missing_keys}. "
f"No values provided for keys: {empty_keys}. "
f"Provided join_key_values: {list(join_key_values.keys())}"
)
# Convert the column-oriented table_entity_values into row-wise data.
keys = list(table_entity_values.keys())
# Each row is a tuple of ValueProto objects corresponding to the join keys.
rowise = list(enumerate(zip(*table_entity_values.values())))
# If there are no rows, return empty tuples.
if not rowise:
return (), (), 0
# Sort rowise so that rows with the same join key values are adjacent.
rowise.sort(key=lambda row: tuple(getattr(x, x.WhichOneof("val")) for x in row[1]))
# Group rows by their composite join key value.
groups = [
(dict(zip(keys, key_tuple)), [idx for idx, _ in group])
for key_tuple, group in itertools.groupby(rowise, key=lambda row: row[1])
]
# If no groups were formed (should not happen for valid input), return empty tuples.
if not groups:
return (), (), 0
# Unpack the unique entities and their original row indexes.
unique_entities, indexes = tuple(zip(*groups))
return unique_entities, indexes, len(rowise)
def _get_unique_entities_from_values(
table_entity_values: Dict[str, List[ValueProto]],
) -> Tuple[Tuple[Dict[str, ValueProto], ...], Tuple[List[int], ...], int]:
"""Return the set of unique composite Entities for a Feature View and the indexes at which they appear.
This method allows us to query the OnlineStore for data we need only once
rather than requesting and processing data for the same combination of
Entities multiple times.
"""
keys = table_entity_values.keys()
# Sort the rowise data to allow for grouping but keep original index. This lambda is
# sufficient as Entity types cannot be complex (ie. lists).
rowise = list(enumerate(zip(*table_entity_values.values())))
rowise.sort(key=lambda row: tuple(getattr(x, x.WhichOneof("val")) for x in row[1]))
# Identify unique entities and the indexes at which they occur.
unique_entities: Tuple[Dict[str, ValueProto], ...]
indexes: Tuple[List[int], ...]
unique_entities, indexes = tuple(
zip(
*[
(dict(zip(keys, k)), [_[0] for _ in g])
for k, g in itertools.groupby(rowise, key=lambda x: x[1])
]
)
)
return unique_entities, indexes, len(rowise)
def _drop_unneeded_columns(
online_features_response: GetOnlineFeaturesResponse,
requested_result_row_names: Set[str],
):
"""
Unneeded feature values such as request data and unrequested input feature views will
be removed from 'online_features_response'.
Args:
online_features_response: Protobuf object to populate
requested_result_row_names: Fields from 'result_rows' that have been requested, and
therefore should not be dropped.
"""
# Drop values that aren't needed
unneeded_feature_indices = [
idx
for idx, val in enumerate(online_features_response.metadata.feature_names.val)
if val not in requested_result_row_names
]
for idx in reversed(unneeded_feature_indices):
del online_features_response.metadata.feature_names.val[idx]
del online_features_response.results[idx]
def _populate_result_rows_from_columnar(
online_features_response: GetOnlineFeaturesResponse,
data: Dict[str, List[ValueProto]],
):
timestamp = Timestamp() # Only initialize this timestamp once.
# Add more values to the existing result rows
for feature_name, feature_values in data.items():
online_features_response.metadata.feature_names.val.append(feature_name)
online_features_response.results.append(
GetOnlineFeaturesResponse.FeatureVector(
values=feature_values,
statuses=[FieldStatus.PRESENT] * len(feature_values),
event_timestamps=[timestamp] * len(feature_values),
)
)
def get_needed_request_data(
grouped_odfv_refs: List[Tuple["OnDemandFeatureView", List[str]]],
) -> Set[str]:
needed_request_data: Set[str] = set()
for odfv, _ in grouped_odfv_refs:
odfv_request_data_schema = odfv.get_request_data_schema()
# if odfv.write_to_online_store, we should not pass in the request data
needed_request_data.update(odfv_request_data_schema.keys())
return needed_request_data
def ensure_request_data_values_exist(
needed_request_data: Set[str],
request_data_features: Dict[str, List[Any]],
):
if len(needed_request_data) != len(request_data_features.keys()):
missing_features = [
x for x in needed_request_data if x not in request_data_features
]
raise RequestDataNotFoundInEntityRowsException(feature_names=missing_features)
def _convert_entity_key_to_proto_to_dict(
entity_key_vals: List[EntityKeyProto],
) -> Dict[str, List[ValueProto]]:
entity_dict: Dict[str, List[ValueProto]] = {}
for entity_key_val in entity_key_vals:
if entity_key_val is not None:
for join_key, entity_value in zip(
entity_key_val.join_keys, entity_key_val.entity_values
):
if join_key not in entity_dict:
entity_dict[join_key] = []
# python_entity_value = _proto_value_to_value_type(entity_value)
entity_dict[join_key].append(entity_value)
return entity_dict
def _get_features(
registry,
project,
features: Union[List[str], "FeatureService"],
allow_cache: bool = False,
) -> List[str]:
from feast.feature_service import FeatureService
_features = features
if not _features:
raise ValueError("No features specified for retrieval")
_feature_refs = []
if isinstance(_features, FeatureService):
# Create cache key for feature service resolution
cache_key = f"{_features.name}:{project}:{hash(tuple(str(fv) for fv in _features.feature_view_projections))}"
# Check cache first if caching is enabled and available
if allow_cache and hasattr(registry, "_feature_service_cache"):
if cache_key in registry._feature_service_cache:
return registry._feature_service_cache[cache_key]
# Resolve feature service from registry
feature_service_from_registry = registry.get_feature_service(
_features.name, project, allow_cache
)
if feature_service_from_registry != _features:
warnings.warn(
"The FeatureService object that has been passed in as an argument is "
"inconsistent with the version from the registry. Potentially a newer version "
"of the FeatureService has been applied to the registry."
)
# Build feature reference list
for projection in feature_service_from_registry.feature_view_projections:
_feature_refs.extend(
[f"{projection.name_to_use()}:{f.name}" for f in projection.features]
)
# Cache the result if caching is enabled and available
if allow_cache and hasattr(registry, "_feature_service_cache"):
registry._feature_service_cache[cache_key] = _feature_refs
else:
assert isinstance(_features, list)
_feature_refs = _features
return _feature_refs
def _list_feature_views(
registry,
project,
allow_cache: bool = False,
hide_dummy_entity: bool = True,
tags: Optional[dict[str, str]] = None,
) -> List["FeatureView"]:
from feast.feature_view import DUMMY_ENTITY_NAME
feature_views = []
for fv in registry.list_feature_views(project, allow_cache=allow_cache, tags=tags):
if hide_dummy_entity and fv.entities and fv.entities[0] == DUMMY_ENTITY_NAME:
fv.entities = []
fv.entity_columns = []
feature_views.append(fv)
return feature_views
def _get_feature_views_to_use(
registry: "BaseRegistry",
project,
features: Union[List[str], "FeatureService"],
allow_cache=False,
hide_dummy_entity: bool = True,
) -> Tuple[List["FeatureView"], List["OnDemandFeatureView"]]:
from feast.feature_service import FeatureService
from feast.feature_view import DUMMY_ENTITY_NAME
from feast.on_demand_feature_view import OnDemandFeatureView
if isinstance(features, FeatureService):
feature_views = [
(projection.name, None, projection)
for projection in features.feature_view_projections
]
else:
assert features is not None
# Parse version-qualified refs: 'fv@v2:feat' -> ('fv', 2, None)
parsed = []
seen = set()
for feature in features:
fv_name, version_num, _ = _parse_feature_ref(feature)
key = (fv_name, version_num)
if key not in seen:
seen.add(key)
parsed.append((fv_name, version_num, None))
feature_views = parsed # type: ignore[assignment]
fvs_to_use, od_fvs_to_use = [], []
for name, version_num, projection in feature_views:
if version_num is not None:
if not getattr(registry, "enable_online_versioning", False):
raise ValueError(
f"Version-qualified ref '{name}@v{version_num}' not supported: "
f"online versioning is disabled. Set 'enable_online_feature_view_versioning: true' "
f"under 'registry' in feature_store.yaml."
)
# Version-qualified reference: look up the specific version snapshot
try:
fv = registry.get_feature_view_by_version(
name, project, version_num, allow_cache
)
except NotImplementedError:
# Fall back for v0 on registries that don't implement versioned lookup
if version_num == 0:
fv = registry.get_any_feature_view(name, project, allow_cache)
else:
raise
# Set version_tag on the projection so name_to_use() returns versioned key
if hasattr(fv, "projection") and fv.projection is not None:
fv.projection.version_tag = version_num
else:
fv = registry.get_any_feature_view(name, project, allow_cache)
if hasattr(fv, "enabled") and not fv.enabled:
raise ValueError(
f"Feature view '{name}' is disabled and cannot serve features. "
f"Enable it with `feast feature-views enable {name}` or set enabled=True."
)
# Enforce lifecycle state gate: only serve if state is AVAILABLE_ONLINE
# or STATE_UNSPECIFIED (backward compat for pre-state feature views).
if hasattr(fv, "state"):
from feast.feature_view import FeatureViewState
if isinstance(fv.state, FeatureViewState) and fv.state not in (
FeatureViewState.STATE_UNSPECIFIED,
FeatureViewState.AVAILABLE_ONLINE,
):
raise ValueError(
f"Feature view '{name}' is in state '{fv.state.name}' "
f"and cannot serve features. Only AVAILABLE_ONLINE feature views can serve."
)
if isinstance(fv, OnDemandFeatureView):
od_fvs_to_use.append(
fv.with_projection(copy.copy(projection)) if projection else fv
)
for source_projection in fv.source_feature_view_projections.values():
source_fv = registry.get_any_feature_view(
source_projection.name, project, allow_cache
)
# TODO better way to handler dummy entities
if (
hide_dummy_entity
and source_fv.entities # type: ignore[attr-defined]
and source_fv.entities[0] == DUMMY_ENTITY_NAME # type: ignore[attr-defined]
):
source_fv.entities = [] # type: ignore[attr-defined]
source_fv.entity_columns = [] # type: ignore[attr-defined]
if source_fv not in fvs_to_use:
fvs_to_use.append(
source_fv.with_projection(copy.copy(source_projection))
)
else:
if (
hide_dummy_entity
and fv.entities # type: ignore[attr-defined]
and fv.entities[0] == DUMMY_ENTITY_NAME # type: ignore[attr-defined]
):
fv.entities = [] # type: ignore[attr-defined]
fv.entity_columns = [] # type: ignore[attr-defined]
if projection:
fv = fv.with_projection(copy.copy(projection))
if version_num is not None:
fv.projection.version_tag = version_num
fvs_to_use.append(fv)
return (fvs_to_use, od_fvs_to_use)
def _get_online_request_context(
registry,
project,
features: Union[List[str], "FeatureService"],
full_feature_names: bool,
):
from feast.feature_view import DUMMY_ENTITY_NAME
_feature_refs = _get_features(registry, project, features, allow_cache=True)
(
requested_feature_views,
requested_on_demand_feature_views,
) = _get_feature_views_to_use(
registry=registry,
project=project,
features=features,
allow_cache=True,
hide_dummy_entity=False,
)
(
entity_name_to_join_key_map,
entity_type_map,
join_keys_set,
) = _get_entity_maps(registry, project, requested_feature_views)
_validate_feature_refs(_feature_refs, full_feature_names)
(
grouped_refs,
grouped_odfv_refs,
) = _group_feature_refs(
_feature_refs,
requested_feature_views,
requested_on_demand_feature_views,
)
# Build expected result names, including version tag when present so
# multi-version queries (e.g. fv@v1:feat, fv@v2:feat) match the response.
requested_result_row_names = set()
for feat_ref in _feature_refs:
fv_name, version_num, feature_name = _parse_feature_ref(feat_ref)
if full_feature_names:
if version_num is not None:
requested_result_row_names.add(
f"{fv_name}@v{version_num}__{feature_name}"
)
else:
requested_result_row_names.add(f"{fv_name}__{feature_name}")
else:
requested_result_row_names.add(feature_name)
feature_views = list(view for view, _ in grouped_refs)
needed_request_data = get_needed_request_data(grouped_odfv_refs)
entityless_case = DUMMY_ENTITY_NAME in [
entity_name
for feature_view in feature_views
for entity_name in (feature_view.entities or [])
]
return (
_feature_refs,
requested_on_demand_feature_views,
entity_name_to_join_key_map,
entity_type_map,
join_keys_set,
grouped_refs,
requested_result_row_names,
needed_request_data,
entityless_case,
)
_feature_resolution_cache: Dict[tuple, tuple] = {}
_feature_resolution_cache_lock = threading.Lock()
_feature_resolution_registry_ts: Optional[datetime] = None
_logger = logging.getLogger(__name__)
def _get_cached_request_context(
registry,
project: str,
features: Union[List[str], "FeatureService"],
full_feature_names: bool,
):
"""Return the output of _get_online_request_context plus resolved ODFV
entity join keys, using a cache that is invalidated whenever the
registry refreshes or its cache TTL expires."""
from feast.feature_service import FeatureService as _FS
global _feature_resolution_cache, _feature_resolution_registry_ts
registry_ts = getattr(registry, "cached_registry_proto_created", None)
if isinstance(features, _FS):
features_key: tuple = ("__fs__", features.name)
else:
features_key = tuple(features)
cache_key = (features_key, project, full_feature_names)
is_cache_valid = getattr(registry, "is_cache_valid", None)
registry_cache_valid = is_cache_valid() if callable(is_cache_valid) else False
if registry_ts is not None and registry_cache_valid:
with _feature_resolution_cache_lock:
if registry_ts != _feature_resolution_registry_ts:
_feature_resolution_cache.clear()
_feature_resolution_registry_ts = registry_ts
_logger.debug("Feature resolution cache cleared (registry refreshed)")
else:
cached = _feature_resolution_cache.get(cache_key)
if cached is not None:
return cached
ctx = _get_online_request_context(registry, project, features, full_feature_names)
(
feature_refs,
requested_on_demand_feature_views,
entity_name_to_join_key_map,
entity_type_map,
join_keys_set,
grouped_refs,
requested_result_row_names,
needed_request_data,
entityless_case,
) = ctx
odfv_join_keys: set = set()
for on_demand_feature_view in requested_on_demand_feature_views:
entities_for_odfv = getattr(on_demand_feature_view, "entities", [])
if len(entities_for_odfv) > 0 and isinstance(entities_for_odfv[0], str):
entities_for_odfv = [
registry.get_entity(entity_name, project, allow_cache=True)
for entity_name in entities_for_odfv
]
odfv_join_keys.update(entities_for_odfv)
result = (
feature_refs,
requested_on_demand_feature_views,
entity_name_to_join_key_map,
entity_type_map,
join_keys_set | odfv_join_keys,
grouped_refs,
frozenset(requested_result_row_names),
needed_request_data,
entityless_case,
)
registry_ts_after = getattr(registry, "cached_registry_proto_created", None)
if registry_ts_after is not None:
with _feature_resolution_cache_lock:
if registry_ts_after != _feature_resolution_registry_ts:
_feature_resolution_cache.clear()
_feature_resolution_registry_ts = registry_ts_after
_feature_resolution_cache[cache_key] = result
return result
def _prepare_entities_to_read_from_online_store(
registry,
project,
features: Union[List[str], "FeatureService"],
entity_values: Mapping[
str, Union[Sequence[Any], Sequence[ValueProto], RepeatedValueProto]
],
full_feature_names: bool = False,
native_entity_values: bool = True,
):
from feast.feature_view import DUMMY_ENTITY, DUMMY_ENTITY_ID, DUMMY_ENTITY_VAL
(
feature_refs,
requested_on_demand_feature_views,
entity_name_to_join_key_map,
entity_type_map,
join_keys_set,
grouped_refs,
requested_result_row_names_frozen,
needed_request_data,
entityless_case,
) = _get_cached_request_context(registry, project, features, full_feature_names)
# Mutable copy downstream code adds join keys to this set.
requested_result_row_names = set(requested_result_row_names_frozen)
# Extract Sequence from RepeatedValue Protobuf.
entity_value_lists: Dict[str, Union[List[Any], List[ValueProto]]] = {
k: list(v) if isinstance(v, Sequence) else list(v.val)
for k, v in entity_values.items()
}
entity_proto_values: Dict[str, List[ValueProto]]
if native_entity_values:
# Convert values to Protobuf once.
entity_proto_values = {
k: python_values_to_proto_values(
v, entity_type_map.get(k, ValueType.UNKNOWN)
)
for k, v in entity_value_lists.items()
}
else:
entity_proto_values = entity_value_lists
num_rows = _validate_entity_values(entity_proto_values)
join_key_values: Dict[str, List[ValueProto]] = {}
request_data_features: Dict[str, List[ValueProto]] = {}
# Entity rows may be either entities or request data.
for join_key_or_entity_name, values in entity_proto_values.items():
# Found request data
if join_key_or_entity_name in needed_request_data:
request_data_features[join_key_or_entity_name] = values
elif join_key_or_entity_name in join_keys_set:
# It's a join key
join_key = join_key_or_entity_name
requested_result_row_names.add(join_key)
join_key_values[join_key] = values
elif join_key_or_entity_name in entity_name_to_join_key_map:
# It's an entity name (deprecated)
join_key = entity_name_to_join_key_map[join_key_or_entity_name]
warnings.warn("Using entity name is deprecated. Use join_key instead.")
requested_result_row_names.add(join_key)
join_key_values[join_key] = values
else:
# Key is not recognized (likely a feature value), so we skip it.
continue # Or handle accordingly
ensure_request_data_values_exist(needed_request_data, request_data_features)
# Populate online features response proto with join keys and request data features
online_features_response = GetOnlineFeaturesResponse(results=[])
_populate_result_rows_from_columnar(
online_features_response=online_features_response,
data=dict(**join_key_values, **request_data_features),
)
# Add the Entityless case after populating result rows to avoid having to remove
# it later.
if entityless_case:
join_key_values[DUMMY_ENTITY_ID] = python_values_to_proto_values(
[DUMMY_ENTITY_VAL] * num_rows, DUMMY_ENTITY.value_type
)
return (
join_key_values,
grouped_refs,
entity_name_to_join_key_map,
requested_on_demand_feature_views,
feature_refs,
requested_result_row_names,
online_features_response,
)
def _get_entity_key_protos(
entity_rows: Iterable[Mapping[str, ValueProto]],
) -> List[EntityKeyProto]:
# Instantiate one EntityKeyProto per Entity.
entity_key_protos = [
EntityKeyProto(join_keys=row.keys(), entity_values=row.values())
for row in entity_rows
]
return entity_key_protos
def _populate_response_from_feature_data(
requested_features: List[str],
read_rows: List[Tuple[Optional[datetime], Optional[Dict[str, ValueProto]]]],
indexes: Iterable[List[int]],
online_features_response: GetOnlineFeaturesResponse,
full_feature_names: bool,
table: "FeatureView",
output_len: int,
include_feature_view_version_metadata: bool = False,
):
"""Populate the GetOnlineFeaturesResponse from raw online_read rows.
Converts raw rows from the OnlineStore into protobuf FeatureVectors and
appends them to the response. This method assumes that ``online_read``
returns data for each unique entity in the same order as ``indexes``.
Args:
requested_features: The names of the features to extract from
each row. Determines the order of FeatureVectors in the response.
read_rows: Raw output from ``OnlineStore.online_read`` a list of
``(event_timestamp, feature_dict)`` tuples, one per unique entity.
``feature_dict`` may be ``None`` when the entity is not found.
indexes: A tuple of lists that maps each unique entity (by position
in ``read_rows``) to one or more output positions in the response.
Used to fan-out deduplicated reads back to the original request rows.
online_features_response: The protobuf response object to populate.
full_feature_names: If True, feature names are prefixed with the
feature view name (e.g. ``"driver_fv__trips_today"``).
table: The FeatureView that ``read_rows`` was retrieved from.
output_len: Total number of result rows in the response.
include_feature_view_version_metadata: If True, version metadata
for the feature view is added to the response.
"""
n_features = len(requested_features)
table_name = table.projection.name_to_use()
clean_table_name = table.projection.name_alias or table.projection.name
feature_refs = [
f"{table_name}__{fn}" if full_feature_names else fn for fn in requested_features
]
online_features_response.metadata.feature_names.val.extend(feature_refs)
if include_feature_view_version_metadata:
existing_names = [
fvm.name for fvm in online_features_response.metadata.feature_view_metadata
]
if clean_table_name not in existing_names:
fv_metadata = online_features_response.metadata.feature_view_metadata.add()
fv_metadata.name = clean_table_name
fv_metadata.version = getattr(table, "current_version_number", 0) or 0
null_value = ValueProto()
null_ts = Timestamp()
PRESENT = FieldStatus.PRESENT
NOT_FOUND = FieldStatus.NOT_FOUND
row_ts_protos = []
for row_ts, _ in read_rows:
ts = Timestamp()
if row_ts is not None:
ts.FromDatetime(row_ts)
row_ts_protos.append(ts)
ts_template = [null_ts] * output_len
indexes_tuple = tuple(indexes)
for row_idx, destinations in enumerate(indexes_tuple):
ts = row_ts_protos[row_idx]
for out_idx in destinations:
ts_template[out_idx] = ts
feat_values = [[null_value] * output_len for _ in range(n_features)]
feat_statuses = [[NOT_FOUND] * output_len for _ in range(n_features)]
feat_idx_map = {name: i for i, name in enumerate(requested_features)}
for row_idx, destinations in enumerate(indexes_tuple):
_, feature_data = read_rows[row_idx]
if feature_data is None:
continue
for feat_name, feat_val in feature_data.items():
f_idx = feat_idx_map.get(feat_name)
if f_idx is not None:
for out_idx in destinations:
feat_values[f_idx][out_idx] = feat_val
feat_statuses[f_idx][out_idx] = PRESENT
try:
from feast.metrics import track_feature_statuses
_present = sum(s == PRESENT for row in feat_statuses for s in row)
_not_found = (n_features * output_len) - _present
track_feature_statuses(table.name, _present, _not_found)
except Exception:
pass
for f_idx in range(n_features):
online_features_response.results.append(
GetOnlineFeaturesResponse.FeatureVector(
values=feat_values[f_idx],
statuses=feat_statuses[f_idx],
event_timestamps=list(ts_template),
)
)
def has_all_tags(
object_tags: dict[str, str], requested_tags: Optional[dict[str, str]] = None
) -> bool:
if requested_tags is None:
return True
return all(object_tags.get(key, None) == val for key, val in requested_tags.items())
def tags_list_to_dict(
tags_list: Optional[list[str]] = None,
) -> Optional[dict[str, str]]:
if not tags_list:
return None
tags_dict: dict[str, str] = {}
for tags_str in tags_list:
tags_dict.update(tags_str_to_dict(tags_str))
return tags_dict
def tags_str_to_dict(tags: str = "") -> dict[str, str]:
tags_list = tags.strip().strip("()").replace('"', "").replace("'", "").split(",")
return {
key.strip(): value.strip()
for key, value in dict(
cast(tuple[str, str], tag.split(":", 1)) for tag in tags_list if ":" in tag
).items()
}
def _utc_now() -> datetime:
return datetime.now(tz=timezone.utc)
def _serialize_vector_to_float_list(vector: List[float]) -> ValueProto:
return ValueProto(float_list_val=FloatListProto(val=vector))
def _build_retrieve_online_document_record(
entity_key: Union[str, bytes],
feature_value: Union[str, bytes],
vector_value: Union[str, List[float]],
distance_value: float,
event_timestamp: datetime,
entity_key_serialization_version: int,
) -> Tuple[
Optional[datetime],
Optional[EntityKeyProto],
Optional[ValueProto],
Optional[ValueProto],
Optional[ValueProto],
]:
if entity_key_serialization_version < 3:
entity_key_proto = None
else:
if isinstance(entity_key, str):
entity_key_proto_bin = entity_key.encode("utf-8")
else:
entity_key_proto_bin = entity_key
entity_key_proto = deserialize_entity_key(
entity_key_proto_bin,
entity_key_serialization_version=entity_key_serialization_version,
)
feature_value_proto = ValueProto()
if isinstance(feature_value, str):
feature_value_proto.ParseFromString(feature_value.encode("utf-8"))
else:
feature_value_proto.ParseFromString(feature_value)
if isinstance(vector_value, str):
vector_value_proto = ValueProto(string_val=vector_value)
else:
vector_value_proto = ValueProto(float_list_val=FloatListProto(val=vector_value))
distance_value_proto = ValueProto(float_val=distance_value)
return (
event_timestamp,
entity_key_proto,
feature_value_proto,
vector_value_proto,
distance_value_proto,
)
def _get_feature_view_vector_field_metadata(
feature_view,
) -> Optional[Field]:
vector_fields = [field for field in feature_view.schema if field.vector_index]
if len(vector_fields) > 1:
raise ValueError(
f"Feature view {feature_view.name} has multiple vector fields. Only one vector field per feature view is supported."
)
if not vector_fields:
return None
return vector_fields[0]