rspsp: Spatio Spectral Analysis in R

Lattice data (meaning measurements taken on a raster) are abundantly available in environmental sciences, geology, astrophysics, and surface imagery. We developed this package to offer tools to inspect the spectral properties of such data.

Inspecting the spectral properties of lattice data can help get supporting evidence for modeling assumptions and get early results for research questions, such as “has the covariance structure between two samples significantly changed?”.

Features

For now we offer tools for: - Sampling data - Inspecting the spectra - Hypothesis testing on the spectra

Installation

We’ve submitted to CRAN but are not listed as of yet. In the mean time you will have to install it from GitHub

library(devtools)
devtools::install_github("https://github.com/Carsten134/rspsp")

Usage Example

Say you want to inspect whether the covariane structure between two lattice data samples differ. For this you can use our test.spectral.

# simulate 20 x 20 MA grid under H1
K0 <- MA_coef_col(.7)
K1 <- MA_coef_row(.7)
x <- gridMA(20, 20, K0)
y <- gridMA(20, 20, K1)

# apply test statistic
test.spectral(x, y, 100, .05)

please see the Getting Started Guide for more detail.