abacusai.monitor_drift_and_distributions

Classes

EmbeddingFeatureDriftDistribution

Feature distribution for embeddings

ForecastingMonitorSummary

Forecasting Monitor Summary of the latest version of the data.

AbstractApiClass

MonitorDriftAndDistributions

Summary of important model monitoring statistics for features available in a model monitoring instance

Module Contents

class abacusai.monitor_drift_and_distributions.EmbeddingFeatureDriftDistribution(client, distance=None, jsDistance=None, wsDistance=None, ksStatistic=None, psi=None, csi=None, chiSquare=None, averageDrift={})

Bases: abacusai.return_class.AbstractApiClass

Feature distribution for embeddings

Parameters:

  • client (ApiClient) – An authenticated API Client instance

  • distance (list) – Histogram data of KL divergences between the training distribution and the range of values in the specified window.

  • jsDistance (list) – Histogram data of JS divergence between the training distribution and the range of values in the specified window.

  • wsDistance (list) – Histogram data of Wasserstein distance between the training distribution and the range of values in the specified window.

  • ksStatistic (list) – Histogram data of Kolmogorov-Smirnov statistic computed between the training distribution and the range of values in the specified window.

  • psi (list) – Histogram data of Population stability index computed between the training distribution and the range of values in the specified window.

  • csi (list) – Histogram data of Characteristic Stability Index computed between the training distribution and the range of values in the specified window.

  • chiSquare (list) – Histogram data of Chi-square statistic computed between the training distribution and the range of values in the specified window.

  • averageDrift (DriftTypesValue) – Average drift embedding for each type of drift

__repr__()

Return repr(self).

to_dict()

Get a dict representation of the parameters in this class

Returns:

The dict value representation of the class parameters

Return type:

dict

class abacusai.monitor_drift_and_distributions.ForecastingMonitorSummary(client, predictionTimestampCol=None, predictionTargetCol=None, trainingTimestampCol=None, trainingTargetCol=None, predictionItemId=None, trainingItemId=None, forecastFrequency=None, trainingTargetAcrossTime={}, predictionTargetAcrossTime={}, actualsHistogram={}, predictionsHistogram={}, trainHistoryData={}, predictHistoryData={}, targetDrift={}, historyDrift={})

Bases: abacusai.return_class.AbstractApiClass

Forecasting Monitor Summary of the latest version of the data.

Parameters:

  • client (ApiClient) – An authenticated API Client instance

  • predictionTimestampCol (str) – Feature in the data that represents the timestamp column.

  • predictionTargetCol (str) – Feature in the data that represents the target.

  • trainingTimestampCol (str) – Feature in the data that represents the timestamp column.

  • trainingTargetCol (str) – Feature in the data that represents the target.

  • predictionItemId (str) – Feature in the data that represents the item id.

  • trainingItemId (str) – Feature in the data that represents the item id.

  • forecastFrequency (str) – Frequency of data, could be hourly, daily, weekly, monthly, quarterly or yearly.

  • trainingTargetAcrossTime (ForecastingAnalysisGraphData) – Data showing average, p10, p90, median sales across time

  • predictionTargetAcrossTime (ForecastingAnalysisGraphData) – Data showing average, p10, p90, median sales across time

  • actualsHistogram (ForecastingAnalysisGraphData) – Data showing actuals histogram

  • predictionsHistogram (ForecastingAnalysisGraphData) – Data showing predictions histogram

  • trainHistoryData (ForecastingAnalysisGraphData) – Data showing length of history distribution

  • predictHistoryData (ForecastingAnalysisGraphData) – Data showing length of history distribution

  • targetDrift (FeatureDriftRecord) – Data showing drift of the target for all drift types: distance (KL divergence), js_distance, ws_distance, ks_statistic, psi, csi, chi_square

  • historyDrift (FeatureDriftRecord) – Data showing drift of the history for all drift types: distance (KL divergence), js_distance, ws_distance, ks_statistic, psi, csi, chi_square

__repr__()

Return repr(self).

to_dict()

Get a dict representation of the parameters in this class

Returns:

The dict value representation of the class parameters

Return type:

dict

class abacusai.monitor_drift_and_distributions.AbstractApiClass(client, id)
__eq__(other)

Return self==value.

_get_attribute_as_dict(attribute)
class abacusai.monitor_drift_and_distributions.MonitorDriftAndDistributions(client, featureDrifts=None, featureDistributions=None, nestedDrifts=None, forecastingMonitorSummary={}, embeddingsDistribution={})

Bases: abacusai.return_class.AbstractApiClass

Summary of important model monitoring statistics for features available in a model monitoring instance

Parameters:

  • client (ApiClient) – An authenticated API Client instance

  • featureDrifts (list[dict]) – A list of dicts of eligible feature names and corresponding overall feature drift measures.

  • featureDistributions (list[dict]) – A list of dicts of feature names and corresponding feature distributions.

  • nestedDrifts (list[dict]) – A list of dicts of nested feature names and corresponding overall feature drift measures.

  • forecastingMonitorSummary (ForecastingMonitorSummary) – Summary of important model monitoring statistics for features available in a model monitoring instance

  • embeddingsDistribution (EmbeddingFeatureDriftDistribution) – Summary of important model monitoring statistics for features available in a model monitoring instance

__repr__()

Return repr(self).

to_dict()

Get a dict representation of the parameters in this class

Returns:

The dict value representation of the class parameters

Return type:

dict