DistMat3D-class: * DistMat3D class

In hsdar: Manage, Analyse and Simulate Hyperspectral Data

* DistMat3D class

Description

Class to store effectively (large) distance matrices (up to 3D), which can be interpreted as a stack of traditional 2-D distance matrices. Therefore, the first two dimensions are of equal length and usually describe the wavelength in hsdar. This third dimension is normally the number of samples or pixels. In hsdar, objects of class DistMat3D are used e.g., to store nri-values. In this case, the first and second dimensions store the information which band #1 is substraced by which band #2, respectively. The third dimension is the sample. Since it usually does not matter if band #1 is substracted from band #2 or vice versa, the nri-matrix would contain the same absolute values on both triangles (as 2-D distance matrices would do). Therefore, hsdar defines and uses the class DistMat3D in which only one triangle is stored and memory demand is considerably reduced.

Details

S4-class with 3 slots:

• values: Numerical vector containing distance values

• ncol: Number of columns in the 3D-matrix. Number of columns equals always the number of rows

• nlyr: Number of layers in the 3D-matrix

The data in the values slot is organized as follows: The first value is the distance at band #1 and band #2 for sample number #1, the second one is for band #1 and band #3 (sample #1) and so forth. Methods to create objects of class DistMat3D for matrix and array objects exist. Additionally, methods to apply functions to the values exist.

Note

See figure in hsdar-package for an overview of classes in hsdar.

Author(s)

Lukas Lehnert

See Also

distMat3D, apply.DistMat3D

hsdar documentation built on March 18, 2022, 6:35 p.m.