Skip to content

Feature Monitoring Result#

[source]

FeatureMonitoringResult#

hsfs.core.feature_monitoring_result.FeatureMonitoringResult(
    feature_store_id,
    execution_id,
    monitoring_time,
    config_id,
    feature_name,
    difference=None,
    shift_detected=False,
    detection_statistics_id=None,
    reference_statistics_id=None,
    empty_detection_window=False,
    empty_reference_window=False,
    specific_value=None,
    raised_exception=False,
    detection_statistics=None,
    reference_statistics=None,
    id=None,
    href=None,
    **kwargs
)

Retrieval#

[source]

get_history#

FeatureMonitoringConfig.get_history(start_time=None, end_time=None, with_statistics=True)

Fetch the history of the computed statistics and comparison results for this configuration.

Example

# Fetch your feature group or feature view
fg = fs.get_feature_group(name="my_feature_group", version=1)
# Fetch registered config by name
my_monitoring_config = fg.get_feature_monitoring_configs(name="my_monitoring_config")
# Fetch the history of the computed statistics for this configuration
history = my_monitoring_config.get_history(
    start_time="2021-01-01",
    end_time="2021-01-31",
)

Args:

start_time: The start time of the time range to fetch the history for. end_time: The end time of the time range to fetch the history for. with_statistics: Whether to include the computed statistics in the results.

Raises

  • FeatureStoreException: If the feature monitoring config has not been saved.

Properties#

[source]

config_id#

Id of the feature monitoring configuration containing this result.


[source]

detection_statistics#

Feature descriptive statistics computed on the detection window.


[source]

detection_statistics_id#

Id of the feature descriptive statistics computed on the detection window.


[source]

difference#

Difference between detection and reference values. It can be relative or absolute difference, depending on the statistics comparison configuration provided in relative parameter passed to compare_on() when enabling feature monitoring.


[source]

empty_detection_window#

Whether or not the detection window was empty in this feature monitoring run.


[source]

empty_reference_window#

Whether or not the reference window was empty in this feature monitoring run.


[source]

execution_id#

Execution id of the feature monitoring job.


[source]

feature_name#

Name of the feature being monitored.


[source]

feature_store_id#

Id of the Feature Store.


[source]

id#

Id of the feature monitoring result.


[source]

monitoring_time#

Time at which this feature monitoring result was created.


[source]

reference_statistics#

Feature descriptive statistics computed on the reference window.


[source]

reference_statistics_id#

Id of the feature descriptive statistics computed on the reference window.


[source]

shift_detected#

Whether or not shift was detected in the detection window based on the computed statistics and the threshold provided in compare_on() when enabling feature monitoring.


[source]

specific_value#

Specific value used as reference in the statistics comparison.