objective.gen.fun: Objective Value in One-To-One Feature Matching with Balanced...

In matchFeat: One-to-One Feature Matching

Objective Value in One-To-One Feature Matching with Balanced or Unbalanced Data

Description

Calculates the objective value in the multidimensional assignment problem with decomposable costs (MDADC). The dissimilarity function used in this problem is the squared Euclidean distance. The data can be balanced OR unbalanced.

Usage

objective.gen.fun(x, unit, cluster)

Arguments

x	data matrix with feature vectors in rows
unit	vector of unit labels (length should equal number of rows in x)
cluster	vector of cluster labels (length should equal number of rows in x)

Details

See equation (2) in Degras (2022). This function gives the same value as objective.fun when the data are balanced.

Value

Objective value

References

Degras (2022) "Scalable feature matching across large data collections." doi: 10.1080/10618600.2022.2074429

See Also

objective.fun

Examples

data(optdigits)
m <- 10
n <- 100

## Balanced example: both 'objective.fun' and 'objective.gen.fun' work
sigma <- matrix(1:m,m,n)
cluster <- rep(1:m,n)
objective.fun(optdigits$x, sigma, optdigits$unit)
objective.gen.fun(optdigits$x, optdigits$unit, cluster)

## Unbalanced example
idx <- 1:999
objective.gen.fun(optdigits$x[idx,], optdigits$unit[idx], cluster[idx])

matchFeat documentation built on Dec. 16, 2022, 5:11 p.m.