evd (version 2.2-1)

mtransform: GEV Transformations

Description

Transforms to exponential margins under the GEV model.

Usage

mtransform(x, p, inv = FALSE, drp = FALSE)

Arguments

x

A matrix with n rows and d columns, or a vector. In the latter case, if p is a list with the same length as the vector, it is treated as a matrix with one row. If p is not a list, it is treated as a matrix with one

p

A vector of length three or a matrix with n rows and three columns. It can also be a list of length d, in which case each element must be a vector of length three or a matrix with n rows and three columns.

inv

Logical; use the inverse transformation?

drp

Logical; return a vector rather than a single row matrix?. Note that a single column matrix is always returned as a vector.

Value

• A numeric matrix or vector.

Details

Let $x_i$ denote a vector of observations for $i = 1,\ldots,n$. This function implements the transformation $$y_{i} = {1+s_i(x_{i}-a_i)/b_i}_{+}^{-1/s_i}$$ to each column of the matrix x.

The values $(a_i,b_i,s_i)$ are contained in the ith row of the n by 3 matrix p. If p is a vector of length three, the parameters are the same for every $i = 1,\ldots,n$. Alternatively, p can be a list with d elements, in which case the jth element is used to transform the jth column of x.

This function is mainly for internal use. It is used by bivariate and multivariate routines to calculate marginal transformations.