dis: Dissimilarities between row observations of a matrix and a...

In mlesnoff/rnirs: Dimension reduction, Regression and Discrimination for Chemometrics

Dissimilarities between row observations of a matrix and a given vector

Description

Calculation of dissimilarities between the row observations of a data set and a given vector mu. By default, if not specified, vector mu is the column means of the reference (= training) data set.

Usage

dis(Xr, Xu = NULL, mu = NULL, diss = c("euclidean", "mahalanobis", "correlation"),
  sigma = NULL, weights = NULL)

Arguments

Xr	A n x p matrix or data frame of reference (= training) observations.
Xu	A m x p matrix or data frame of new (= test) observations (Xu is not used in the calculation of the median and MAD used for calculating the standardized dissimilarities; see Details).
mu	The vector of length p which is compared to the n rows of Xr and, if not NULL, to the m rows Xu. If mu is NULL (default), mu is set to the column means of Xr.
diss	The type of dissimilarities used. Possible values are "euclidean" (Euclidean distances; default), "mahalanobis" (Mahalanobis distances), or "correlation". Correlation dissimilarities are calculated by sqrt(.5 * (1 - rho)).
sigma	For Mahalanobis distance, the covariance matrix considered. If NULL (default), sigma is set to cov(Xr).
weights	Only for Mahalanobis distance and when sigma = NULL. A vector of length n defining a priori weights to apply to the training observations, for the computation of the covariance matrix. Internally, weights are "normalized" to sum to 1. Default to NULL (weights are set to 1 / n).

Value

A list of outputs, such as:

dr	A data frame of the dissimilarities for Xr.
du	A data frame of the dissimilarities for Xu.

Examples

n <- 8
p <- 6
set.seed(1)
X <- matrix(rnorm(n * p, mean = 10), ncol = p, byrow = TRUE)
y1 <- 100 * rnorm(n)
y2 <- 100 * rnorm(n)
Y <- cbind(y1, y2)
set.seed(NULL)

Xr <- X[1:6, ] ; Yr <- Y[1:6, ]
Xu <- X[7:8, ] ; Yu <- Y[7:8, ]

mu <- NULL
dis(Xr, mu = mu)

dis(Xr, Xu, mu)

fm <- pls(Xr, Yr, ncomp = 3)
dis(fm$Tr, mu = mu, diss = "mahalanobis")

fm <- pls(Xr, Yr, Xu, ncomp = 3)
dis(fm$Tr, fm$Tu, mu, diss = "mahalanobis")

mlesnoff/rnirs documentation built on April 24, 2023, 4:17 a.m.