Implementation:Evidentlyai Evidently Legacy Data Drift Preset

Knowledge Sources	Evidentlyai_Evidently
Domains	Data Quality, Drift Detection, Model Monitoring
Last Updated	2026-02-14 12:00 GMT

Overview

The DataDriftPreset class is a metric preset that bundles dataset-level drift detection, per-column drift analysis, and embeddings drift metrics into a single preset for Evidently reports.

Description

DataDriftPreset extends MetricPreset and generates a list of drift-related metrics via its generate_metrics method. The preset always includes:

DatasetDriftMetric -- detects whether the overall dataset has drifted, using a configurable drift_share threshold (default: 0.5, meaning drift is flagged if more than 50% of columns show drift)
DataDriftTable -- produces a per-column drift analysis table

Both metrics receive the full suite of statistical test configuration parameters, allowing fine-grained control over which statistical tests are used for different column types (categorical, numerical, text) and individual columns.

When embeddings data is present in the data definition, the preset additionally generates EmbeddingsDriftMetric instances for each embedding, using the helper function add_emb_drift_to_reports. Custom drift methods can be specified per embedding via the embeddings_drift_method parameter.

The preset supports extensive configuration of statistical tests:

Global defaults via stattest and stattest_threshold
Type-specific overrides via cat_stattest, num_stattest, text_stattest and their corresponding threshold parameters
Per-column overrides via per_column_stattest and per_column_stattest_threshold dictionaries

Usage

Use this preset when you want comprehensive data drift analysis across your dataset. It is the standard entry point for detecting feature distribution changes between reference and current data, suitable for production monitoring pipelines.

Code Reference

Source Location

Repository: Evidentlyai_Evidently
File: src/evidently/legacy/metric_preset/data_drift.py

Signature

class DataDriftPreset(MetricPreset):
    class Config:
        type_alias = "evidently:metric_preset:DataDriftPreset"

    columns: Optional[List[str]]
    embeddings: Optional[List[str]]
    embeddings_drift_method: Optional[Dict[str, DriftMethod]]
    drift_share: float
    stattest: Optional[PossibleStatTestType]
    cat_stattest: Optional[PossibleStatTestType]
    num_stattest: Optional[PossibleStatTestType]
    text_stattest: Optional[PossibleStatTestType]
    per_column_stattest: Optional[Dict[str, PossibleStatTestType]]
    stattest_threshold: Optional[float]
    cat_stattest_threshold: Optional[float]
    num_stattest_threshold: Optional[float]
    text_stattest_threshold: Optional[float]
    per_column_stattest_threshold: Optional[Dict[str, float]]

    def __init__(self, columns: Optional[List[str]] = None,
                 embeddings: Optional[List[str]] = None,
                 embeddings_drift_method: Optional[Dict[str, DriftMethod]] = None,
                 drift_share: float = 0.5,
                 stattest: Optional[PossibleStatTestType] = None,
                 cat_stattest: Optional[PossibleStatTestType] = None,
                 num_stattest: Optional[PossibleStatTestType] = None,
                 text_stattest: Optional[PossibleStatTestType] = None,
                 per_column_stattest: Optional[Dict[str, PossibleStatTestType]] = None,
                 stattest_threshold: Optional[float] = None,
                 cat_stattest_threshold: Optional[float] = None,
                 num_stattest_threshold: Optional[float] = None,
                 text_stattest_threshold: Optional[float] = None,
                 per_column_stattest_threshold: Optional[Dict[str, float]] = None): ...

    def generate_metrics(self, data_definition: DataDefinition,
                         additional_data: Optional[Dict[str, Any]]) -> List[AnyMetric]: ...

Import

from evidently.legacy.metric_preset.data_drift import DataDriftPreset

I/O Contract

Inputs

Name	Type	Required	Description
columns	Optional[List[str]]	No	Specific columns to analyze for drift. If None, all columns are analyzed.
embeddings	Optional[List[str]]	No	List of embedding names to analyze for drift
embeddings_drift_method	Optional[Dict[str, DriftMethod]]	No	Custom drift detection methods per embedding name
drift_share	float	No	Fraction of drifting columns required to flag dataset-level drift (default: 0.5)
stattest	Optional[PossibleStatTestType]	No	Default statistical test for all column types
cat_stattest	Optional[PossibleStatTestType]	No	Statistical test override for categorical columns
num_stattest	Optional[PossibleStatTestType]	No	Statistical test override for numerical columns
text_stattest	Optional[PossibleStatTestType]	No	Statistical test override for text columns
per_column_stattest	Optional[Dict[str, PossibleStatTestType]]	No	Per-column statistical test overrides
stattest_threshold	Optional[float]	No	Default p-value threshold for drift detection
cat_stattest_threshold	Optional[float]	No	Threshold override for categorical columns
num_stattest_threshold	Optional[float]	No	Threshold override for numerical columns
text_stattest_threshold	Optional[float]	No	Threshold override for text columns
per_column_stattest_threshold	Optional[Dict[str, float]]	No	Per-column threshold overrides

Outputs

Name	Type	Description
metrics	List[AnyMetric]	A list containing DatasetDriftMetric, DataDriftTable, and optionally EmbeddingsDriftMetric instances

Usage Examples

from evidently.legacy.metric_preset.data_drift import DataDriftPreset

# Basic usage with defaults
preset = DataDriftPreset()

# With custom drift share threshold and specific columns
preset = DataDriftPreset(
    columns=["age", "income", "category"],
    drift_share=0.3
)

# With per-type statistical test configuration
preset = DataDriftPreset(
    num_stattest="ks",
    cat_stattest="chi2",
    stattest_threshold=0.05
)

# With embeddings drift detection
preset = DataDriftPreset(
    embeddings=["text_embedding"],
    embeddings_drift_method={"text_embedding": DriftMethod(name="model")}
)

Related Pages

Page Connections

Double-click a node to navigate. Hold to expand connections.

Principle

Implementation

Heuristic

Environment