celphin/Equitable: Equitable Transform

A method of scaling to maintain individuality The equitable transformation maintains systematic differences between instances when the underlying pattern is a separable function in two variables + an addional function of one of these variables. This technique converts “equitably” between comparable instances in a dataset. Applied to a data set, it removes noise superimposed on an underlying signal enabling patterns to be extracted more easily from a 2D data set. Elphinstone, C. and Henry, G. (submitted to Ecosphere, 2019) show the multipurpose uses for the equitable transform. Functions useful for analysing two dimensional data sets are included in this package. A two dimensional data set (d) is transformed via Td<-transformE(d) and a variety of matrices associated with the equitable transform is placed in Td. A signal data set can be made via d<-eg4(2,2) and noise can be added to make a noisy 2D data set (d_noise) in a variety of ways. If Td and Td_noise are the results from running transformE then plotsummary(Td_noise,Td), plotsummary(Td) and plotsummary(Td_noise) summarize and plot information about the two transformations.

install.packages("remotes")
remotes::install_github("celphin/Equitable")