geo-sampling

A collection of geoscience based sampling tools.

Environment

Minimum R version: 3.4

Has been tested on R 3.6, later versions may also work.

Requirements

Required system packages:

• GDAL
• PROJ

Required R packages:

• raster
• clhs
• rgdal
• moments
• rgeos
• BalancedSampling

Installation

If R packages are not installed, geo-sampling will attempt to install required R packages from CRAN.

Otherwise, they can be installed using:

install.packages(c('raster, 'clhs', 'rgdal', 'moments', 'rgeos', 'BalancedSampling'))

There are also some scripts to assist with installing on Gadi (NCI) and Ubuntu 18.04.

Note: at time of writing, cLHS is not available on CRAN. It will have to be downloaded and installed manually. See the installation scripts for details on how to do this.

Gadi:

To install on gadi, run:

./gadi_install.sh

By default, R libraries will be installed to your home directory on the NCI. To install libararies to another directory, set the desired R library path by setting the R_LIBS environment variable before running the installation:

export R_LIBS=/path/to/R/library

Ubuntu 18.04

To install on Ubuntu 18.04, run:

./ubuntu1804_install.sh

The Ubuntu scripts installs required system packages (GDAL, PROJ) and R packages. Installing system packages requires sudo privileges, so this script may request your unix password.

By default, R libraries will be installed to the R user library directory ("$HOME/R/x86_64-pc-linux-gnu-library/3.4/"). If you wish to install to a different directory, you can modify the LIB_PATH variable on line 15 of the installation script.

This script may work for other versions of Ubuntu. It may also work for other Linux distributions, but the names of packages and the package manager may be different and require modification.

How To Use

Configurations are stored in Run_Scripts.R file. So, first make your changes (e.g. change the input pathes to your choice).

Currently the following functionalities are supported:

conditioned Latin Hypercube Sampling (cLHS)

• To extract a handful of points at the target locations,
• To first extract points around given a buffer size and then extract a handful of points,
• The above scenarios with an existing model to weight the inputs.

Comment or uncomment relevant lines in the file Run_Scripts.R according to what functionality you want and then from within the geo-sampling directory, run:

Rscript Run_Scripts.R

Once complete, check the output folder for results.

For more detailed instructions, see: walkthrough

Gadi

When running on the NCI, you also need to load the required modules and set up your R library path before running a job. These can be loaded with the following commands:

module load R/3.6.1
module load gdal/3.0.2
module load proj/6.2.1

The Run_NCI.sh script can assist with this, and can also be used as a template for submitting jobs.