initRandrot-class: Initialised random rotation class

In randRotation: Random Rotation Methods for High Dimensional Data with Batch Structure

Description

List-based S4 class containing all information necessary to generate randomly rotated data with the randrot method. initRandrot and initRandrotW objects are created with the initRandrot method.

initRandrotW is organised as its base class initRandrot, altough some components are changed or added.

Components

The following components are included as list elements:

X

Original (non-transformed) design matrix.

Xhe, Xhe.Y.w, Yd

Pre-multiplied matrix products needed for generation of rotated data (randrot).

coef.h, coef.d

Indices of H0 coefficients (coef.h or "hypothesis coefficients") and indices of all other coefficients (coef.d or "determined coefficients").

cormat

Correlation matrix, see initRandrot.

tcholC

Cholesky decomposition of cormat: cormat = crossprod(tcholC).

rank

Rank of the qr decomposition of (transformed/whitened) X

The following components are changed or added in initRandrotW-class as compared to initRandrot-class:

decomp.list

List containing Xd, Xhe and rank of the transformed/whitened design matrix for each feature, see also X_decomp.

w

Numeric matrix with dimensions features x samples containing component wise square root of the weight matrix, see initRandrot.

Author(s)

Peter Hettegger

randRotation documentation built on April 14, 2021, 6:01 p.m.