PCAs4clust: PCAs4clust

In LPDynR: Land Productivity Dynamics Indicator

PCAs4clust

Description

PCAs4clust runs a two-steps process to prepare the data to be clustered

Usage

PCAs4clust(obj2process = NULL, cumul_var_threshold = 0.9, filename = "", ...)

Arguments

obj2process	SpatRaster object (or its file name). Each layer is one variable
cumul_var_threshold	Numeric. Optional (default = 0.9). Threshold of cumulative variance to select the number of useful PCs
filename	Character. Output filename. Optional
...	Optional. Arguments for prcomp

Details

Firstly, a Principal Component Analysis ('screening PCA') with all the variables in 'obj2process' is run in order to know the optimal number of variables to be used in a subsequent PCA, as well as the most associated variable to those Principal Components (PCs). A threshold of cumulative variance (cumul_var_threshold; default = 0.9) is needed. Secondly, a 'final PCA' is run with the results of the 'screening PCA' (i.e. number of PC axes and their most associated variables). PCAs4clust uses prcomp to run PCAs

Value

SpatRaster object

Author(s)

Xavier Rotllan-Puig

See Also

rm_multicol; prcomp

Examples

dirctry <- paste0(system.file(package='LPDynR'), "/extdata")
variables_noCor <- rm_multicol(dir2process = dirctry,
                               multicol_cutoff = 0.7)
PCAs4clust(obj2process = variables_noCor,
            cumul_var_threshold = 0.9)

LPDynR documentation built on Oct. 16, 2023, 5:06 p.m.