Attribute: Attribute specification

In cleanzr/dblinkR: R interface for dblink

Description

Represents an entity attribute that appears in the data: e.g. 'name of company', 'age', 'zip code', etc.

Usage

Attribute(simFn, distortionPrior)

is.Attribute(x)

CategoricalAttribute(distortionPrior)

is.CategoricalAttribute(x)

Arguments

simFn	a SimFn' object.
distortionPrior	a BetaRV object. Specifies the prior on the distortion probability for the attribute.
x	an R object

Details

An Attribute object contains all of the model parameters associated with an entity attribute. This includes:

• the prior distribution over the distortion probability (likelihood that an entity attribute is distorted in the data), and

• the distance function (parameterizes the distortion distribution).

A named list of Attribute objects is required when initializing the model (see initializeState).

Value

The Attribute constructor returns an Attribute object.

is.Attribute returns TRUE if the argument is an Attribute object and FALSE otherwise.

The CategoricalAttribute constructor returns a CategoricalAttribute object. It is intended for modeling categorical attributes, and uses a a constant similarity function.

is.CategoricalAttribute returns TRUE if the argument is a CategoricalAttribute object and FALSE otherwise.

cleanzr/dblinkR documentation built on June 13, 2021, 4:17 a.m.