Unfolding function

Multidimensional Unfolding

Multidimensional Unfolding with some adaptations for vegetation analysis

Unfolding(A, ENV = NULL, TransAbund = "Gaussian Columns", offset = 0.5, 
weight = "All_1", Constrained = FALSE, 
TransEnv = "Standardize columns", 
InitConfig = "SVD", model = "Ratio", 
condition = "Columns", Algorithm = "SMACOF", 
OptimMethod = "CG", r = 2, maxiter = 100, 
tolerance = 1e-05, lambda = 1, omega = 0, plot = FALSE)

Arguments

• A: The original proximities matrix
• ENV: The matrix of environmental variables
• TransAbund: Initial transformation of the abundances : "None", "Gaussian", "Column Percent", "Gaussian Columns", "Inverse Square Root", "Divide by Column Maximum")
• offset: offset is the quantity added to the zeros of the table
• weight: A matrix of weights for each cell of the table
• Constrained: Should fit a constrained analysis
• TransEnv: Transformation of the environmental variables
• InitConfig: Init configuration for the algorithm
• model: Type of model to be fitted: "Identity", "Ratio", "Interval" or "Ordinal".
• condition: "Matrix", "Columns" to condition to the whole matrix or to each column
• Algorithm: Algorithm to fit the model: "SMACOF", "GD", "Genefold"
• OptimMethod: Optimization method for gradient descent
• r: Dimension of the solution
• maxiter: Maximum number of iterations in the algorithm
• tolerance: Tolerace for the algorithm
• lambda: First penalization parameter
• omega: Second penalization parameter
• plot: Should the results be plotted?

Details

ological data

Returns

An object of class "Unfolding"

References

Ver Articulos

Author(s)

Jose Luis Vicente Villardon

Examples

unf=Unfolding(SpidersSp, ENV=SpidersEnv, model="Ratio", Constrained = FALSE, condition="Matrix")
plot(unf, PlotTol=TRUE, PlotEnv = FALSE)
plot(unf, PlotTol=TRUE, PlotEnv = TRUE)
cbind(unf$QualityVars, unf$Var_Fit)
unf2=Unfolding(SpidersSp, ENV=SpidersEnv, model="Ratio", Constrained = TRUE, condition="Matrix")
plot(unf2, PlotTol=FALSE, PlotEnv = TRUE, mode="s")
cbind(unf2$QualityVars, unf2$Var_Fit)