README.md

Island

This package contains a set of functions to make handling topographically complex island datasets a little less frustrating. It also contains functions to facilitate the parallel processing of geographic data using the R package raster. The package is currently very early in its development; documentation, worked examples and tidy code is currently unfinished.

At its core, the package relies heavily on the package raster. I've simply taken existing functions and packaged them into some useful functions of my own to (hopefully) make the initial stages of analysis that bit easier.

Background

The package was bourne out of my frustration at handling climate data for islands. The complex coastlines of most islands present unique challenges when working with climate data from different products (e.g. temperature from Worldclim and precipitation from CHELSA). Different products would have the odd coastal raster cell present or missing - when running a species distribution model, for example, the model would fail as some climate data was present while other times it was not. To fix this problem (and others) I started developing this package as I'm surely not the first (or last!) person to have experienced it. Over time, I included other functions to split and merge rasters in order to run models in paralell; I've included them here

Installation

To install island:

library(devtools)
install_github("simon-tarr/island")

Example workflow

I shall update this section with a worked example soon. For now though the package could be used in the following way:

  1. Load in your climate data. Use sync_na to ensure that there's climate data for each cell across your study area.
  2. Use split_raster to tile your study area into smaller chunks. These can then be used within foreach or another parallel processing package of your choice.
  3. Stitch your results back together using stitch_raster.
  4. Profit.


simon-tarr/island documentation built on May 6, 2019, 8:05 p.m.