Classifier

class

class Classifier

Bases: ABC, JsonSerde

Categories: dq-classifier

Abstract base class for all data quality classifiers.

A classifier evaluates one or more columns and writes a per-row verdict into a new column. Collect classifiers in a DataQuality action (classifiers=[...]) alongside the operators that act on their verdicts. Concrete classifiers are frozen; build one per rule.

Concrete classifiers share these construction fields (a few fix on_wrong_type / on_wrong_value and do not expose them):

Examples

Classifiers are the labelling half of a check; pass concrete ones to DataQuality(classifiers=[...]):

DataQuality(
    table="checked",
    classifiers=[IsNotNull(["customer_id"]), IsPositive(["amount"])],
    operators=[Enrich()],
)

Methods

method

supported_dtypes

def supported_dtypes() -> frozenset

Returns the set of data types supported by this classifier.