Principal Component Analysis

Author(s)	O.Conrad (c) 2010
Library ID	statistics_grid
Tool ID	8
Version	1.0
Menu	Spatial and Geostatistics \| Grids \| Principal Components

Description

Principal Component Analysis (PCA) for grids. PCA implementation is based on F.Murtagh's code as provided by the StatLib web site.

References

C-code by F.Murtagh

StatLib Web Site

Bahrenberg, G., Giese, E., Nipper, J. (1992): Statistische Methoden in der Geographie 2 - Multivariate Statistik. pp.198-277.

Parameters

	Name	Type	Identifier	Description	Constraints
Input	Grids	grid list, input	`GRIDS`	-	-
Input	Eigen Vectors	table, input, optional	`EIGEN_INPUT`	Use Eigen vectors from this table instead of calculating these from the input grids.	-
Output	Principal Components	grid collection, output	`COLLECTION`	-	-
	Principal Components	grid list, output	`PCA`	-	-
	Eigen Vectors	table, output, optional	`EIGEN`	Store calculated Eigen vectors to this table, e.g. for later use with forward or inverse PCA.	-
Options	Grid System	grid system	`PARAMETERS_GRID_SYSTEM`	-	-
	Output	choice	`OUTPUT`	-	Available Choices: [0] grid collection [1] single grids Default: 0
	Method	choice	`METHOD`	-	Available Choices: [0] correlation matrix [1] variance-covariance matrix [2] sums-of-squares-and-cross-products matrix [3] variance-covariance matrix with normalized features Default: 1
	Number of Components	integer number	`COMPONENTS`	number of first components in the output; set to zero to get all	Minimum: 0 Default: 0

Command Line


Usage: saga_cmd statistics_grid 8 [-GRIDS ] [-OUTPUT ] [-COLLECTION ] [-PCA ] [-EIGEN_INPUT ] [-EIGEN ] [-METHOD ] [-COMPONENTS ]
  -GRIDS:         Grids
	grid list, input
  -OUTPUT:        Output
	choice
	Available Choices:
	[0] grid collection
	[1] single grids
	Default: 0
  -COLLECTION:    Principal Components
	grid collection, output
  -PCA:           Principal Components
	grid list, output
  -EIGEN_INPUT:   Eigen Vectors
	table, input, optional
  -EIGEN:         Eigen Vectors
	table, output, optional
  -METHOD:
                       Method
                       	choice
                       	Available Choices:
                       	[0] correlation matrix
                       	[1] variance-covariance matrix
                       	[2] sums-of-squares-and-cross-products matrix
                       	[3] variance-covariance matrix with normalized features
                       	Default: 1
  -COMPONENTS:    Number of Components
	integer number
	Minimum: 0
	Default: 0