dbstatspackage {dbstats}  R Documentation 
Distancebased statistics (dbstats)
Description
This package contains functions for distancebased prediction methods.
These are methods for prediction where predictor information is coded as a matrix of distances between individuals.
In the currently implemented methods the response is a univariate variable as in the ordinary linear model or in the generalized linear model.
Distances can either be directly input as an distances matrix,
a squared distances matrix, an innerproducts matrix
(see GtoD2
) or computed from observed
explanatory variables.
Notation convention: in distancebased methods we must distinguish observed explanatory variables which we denote by Z or z, from Euclidean coordinates which we denote by X or x. For explanation on the meaning of both terms see the bibliography references below.
Observed explanatory variables z are possibly a mixture of continuous and qualitative explanatory variables or more general quantities.
dbstats does not provide specific functions for computing distances, depending instead on other functions and packages, such as:

dist
in the stats package. 
dist
in the proxy package. When the proxy package is loaded, itsdist
function supersedes the one in the stats package. 
daisy
in the cluster package. Compared to both instances ofdist
above whose input must be numeric variables, the main feature ofdaisy
is its ability to handle other variable types as well (e.g. nominal, ordinal, (a)symmetric binary) even when different types occur in the same data set.Actually the last statement is not hundred percent true: it refers only to the default behaviour of both
dist
functions, whereas thedist
function in the proxy package can evaluate distances between observations with a userprovided function, entered as a parameter, hence it can deal with any type of data. See the examples inpr_DB
.
Functions of dbstats package:
Linear and local linear models with a continuous response:

dblm
for distancebased linear models. 
ldblm
for local distancebased linear models. 
dbplsr
for distancebased partial least squares.
Generalized linear and local generalized linear models with a numeric response:

dbglm
for distancebased generalized linear models. 
ldbglm
for local distancebased generalized linear models.
Details
Package:  dbstats 
Type:  Package 
Version:  2.0.2 
Date:  20240126 
License:  GPL2 
LazyLoad:  yes 
Author(s)
Boj, Eva <evaboj@ub.edu>, Caballe, Adria <adria.caballe@upc.edu>, Delicado, Pedro <pedro.delicado@upc.edu> and Fortiana, Josep <fortiana@ub.edu>
References
Boj E, Caballe, A., Delicado P, Esteve, A., Fortiana J (2016). Global and local distancebased generalized linear models. TEST 25, 170195.
Boj E, Delicado P, Fortiana J (2010). Distancebased local linear regression for functional predictors. Computational Statistics and Data Analysis 54, 429437.
Boj E, Grane A, Fortiana J, Claramunt MM (2007). Implementing PLS for distancebased regression: computational issues. Computational Statistics 22, 237248.
Boj E, Grane A, Fortiana J, Claramunt MM (2007). Selection of predictors in distancebased regression. Communications in Statistics B  Simulation and Computation 36, 8798.
Cuadras CM, Arenas C, Fortiana J (1996). Some computational aspects of a distancebased model for prediction. Communications in Statistics B  Simulation and Computation 25, 593609.
Cuadras C, Arenas C (1990). A distancebased regression model for prediction with mixed data. Communications in Statistics A  Theory and Methods 19, 22612279.
Cuadras CM (1989). Distance analysis in discrimination and classification using both continuous and categorical variables. In: Y. Dodge (ed.), Statistical Data Analysis and Inference. Amsterdam, The Netherlands: NorthHolland Publishing Co., pp. 459473.