README.md

In lindesaysh/MRSea: Spatially adaptive regression splines using SALSA

MRSea

The latest stable version is 1.6 (30/04/2024)

The latest development version is on the development branch.

The MRSea packages allows the fitting of spatially adaptive regression splines using SALSA.

It was developed to examine animal survey data for signs of changes in animal abundance and distribution following marine renewables development. However the methods are suitable for a wide range of applications.

The functions of this package can be used to analyse segmented line transect (alongside the mrds package) or digital aerial data. The package includes functions for fitting spatially adaptive one and 2D smoothers using SALSA and CReSS. Euclidean or Geodesic distances can be used to underpin the smoothed 2D surface and a choice of Gaussian or exponential radial basis functions are available. Non-parametric bootstrapping is available to estimate uncertainty. Several model assessment tools are also available. For models with residual correlation, direct estimation of robust standard errors, given a panel structure, is available.

Recent updates include the reinstatement of natural cubic splines, the Tweedie distribution and a package website with additional materials.

Installation

You can install the latest bugfix release of MRSea from GitHub with:

# install.packages("devtools")
devtools::install_github("lindesaysh/MRSea", ref="stable")

You can install the development version of MRSea from GitHub with:

devtools::install_github("lindesaysh/MRSea", ref="development")

The package may also be downloaded as a .zip or .tar.gz from the latest release

Documentation

There are two “Getting Started” vignettes available with the package:

• Getting Started with MRSea: One dimensional smoothing
• Getting Started with MRSea: Two dimensional smoothing

These are also on the website along with a number of other tutorials.

lindesaysh/MRSea documentation built on May 11, 2024, 11:30 p.m.