Skip to content

fix: Fix STRING type handling in on-demand feature views with RequestSource #5657

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

Closed
Copilot wants to merge 5 commits into from
Closed

fix: Fix STRING type handling in on-demand feature views with RequestSource #5657

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
5 commits
Select commit Hold shift + click to select a range
2f32945
Initial plan
Copilot Oct 11, 2025
7f131c9
Fix STRING type handling in on-demand feature views
Copilot Oct 11, 2025
b9ad729
Improve object dtype handling with value-based type inference
Copilot Oct 11, 2025
1530894
Fix object dtype handling to not intercept array types
Copilot Oct 12, 2025
77540b1
Fix object dtype check to not return early for arrays
Copilot Oct 12, 2025
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
10 changes: 10 additions & 0 deletions sdk/python/feast/type_map.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -176,6 +176,16 @@ def python_type_to_feast_value_type(
item_type = type_map[str(value.dtype)]
return ValueType[item_type.name + "_LIST"]

# Special handling for pandas 'object' dtype - infer from actual value type
# This must come after array handling to avoid intercepting array types
if type_name == "object" and not isinstance(value, (list, np.ndarray)):
if value is not None:
actual_type = type(value).__name__.lower()
if actual_type in type_map:
return type_map[actual_type]
# Default to STRING for object dtype (pandas commonly uses object for strings)
return ValueType.STRING

if isinstance(value, (list, np.ndarray)):
# if the value's type is "ndarray" and we couldn't infer from "value.dtype"
# this is most probably array of "object",
Expand Down
40 changes: 39 additions & 1 deletion sdk/python/tests/unit/test_on_demand_feature_view.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -17,6 +17,7 @@
import pandas as pd
import pytest

from feast.data_source import RequestSource
from feast.feature_view import FeatureView
from feast.field import Field
from feast.infra.offline_stores.file_source import FileSource
Expand All @@ -26,7 +27,7 @@
PythonTransformation,
on_demand_feature_view,
)
from feast.types import Float32
from feast.types import Float32, String


def udf1(features_df: pd.DataFrame) -> pd.DataFrame:
Expand Down Expand Up @@ -418,3 +419,40 @@ def another_transform(features_df: pd.DataFrame) -> pd.DataFrame:

deserialized = OnDemandFeatureView.from_proto(proto)
assert deserialized.name == CUSTOM_FUNCTION_NAME


def test_on_demand_string_features():
"""Test that on-demand feature views work with STRING type inputs from RequestSource."""
# Define a request data source with STRING type fields
input_request = RequestSource(
name="vals_to_add",
schema=[
Field(name="val_to_add", dtype=String),
Field(name="val_to_add_2", dtype=String),
],
)

# Use input data to create new features in Pandas mode
@on_demand_feature_view(
sources=[input_request],
schema=[
Field(name="conv_rate_plus_val1", dtype=String),
Field(name="conv_rate_plus_val2", dtype=String),
],
mode="pandas",
)
def transformed_conv_rate(features_df: pd.DataFrame) -> pd.DataFrame:
df = pd.DataFrame()
df["conv_rate_plus_val1"] = features_df["val_to_add"]
df["conv_rate_plus_val2"] = features_df["val_to_add_2"]
return df

# This should work without raising ValueError about 'object' type
transformed_conv_rate.infer_features()

# Verify the inferred features have the correct type
assert len(transformed_conv_rate.features) == 2
assert transformed_conv_rate.features[0].name == "conv_rate_plus_val1"
assert transformed_conv_rate.features[0].dtype == String
assert transformed_conv_rate.features[1].name == "conv_rate_plus_val2"
assert transformed_conv_rate.features[1].dtype == String
Loading