rcoalescence is a sister project to pycoalescence for running spatially-explicit neutral models for ecological systems. The package is available for macOS and Linux operating systems.
Package migrated from bitbucket on 07/03/2020.
- Simulate spatially-explicit neutral models with complex map configurations.
- Use varying spatial and temporal sampling regimes to mimic real data.
- Uses coalescence methods for high-performance simulations.
- Convenient, minimalistic code structure.
- Obtain a variety of biodiversity metrics including species richness, species' locations and species abundances.
- Make sure that you have installed R > 3.6.0 with Rtools version 3.5 or later.
- Install the necessary R package requirements using
install.packages(c("devtools", "Rcpp", "RSQLite", "dplyr")). - Run
devtools::install_github("thompsonsed/rcoalescence"). This will also download gdal (if it can't be found already).
- Ensure you have R > 3.4.0 installed.
- Install gdal >2.4.0. If you have homebrew installed, this can be done using
brew install gdal2. Check out the brew site for the command to install brew. - Make sure xcode is installed (
xcode-select --install). - Install the necessary R package requirements using
install.packages(c("devtools", "Rcpp", "RSQLite", "dplyr")). - Run
devtools::install_github("thompsonsed/rcoalescence").
- Ensure you have R > 3.4.0 installed, plus a compiler that supports C++14 or later (e.g. gcc or clang).
- Install gdal >2.4.0 from your distribution's repositories (e.g.
sudo apt-get install gdal-devfor Ubuntu), or build from source. - Install the necessary R package requirements using
install.packages(c("devtools", "Rcpp", "RSQLite", "dplyr")). - Run
devtools::install_github("thompsonsed/rcoalescence").
We also suggest a number of other packages contained within the tidyverse collection. All required packages can be installed using install.packages(c("devtools", "Rcpp", "RSQLite", "tidyverse", "rmarkdown", "knitr")).
See the examples by running ?rcoalescence after importing the package into R. The basic process
for a spatial simulation is:
library(rcoalescence)
# Define the folder containing the maps (the examples come with the package)
map_dir <- system.file("sample", package = "rcoalescence")
# The recommended method is to rely on rcoalescence to detect dimensions and offsets of your maps.
# For specifying all dimensions and offsets manually, please see the examples.
# Create a new object to contain the SpatialTreeSimulation
simulation <- SpatialTreeSimulation$new()
# Set the main simulation parameters
simulation$setSimulationParameters(
seed = 1, # random number seed
task = 10, # the job number for file naming
output_directory = "output", # the output directory
min_speciation_rate = 0.001, # the minimum speciation rate
sigma = 2, # the dispersal sigma value
deme = 10, # the number of individuals per cell (multiplied by the map values)
deme_sample = 0.1, # the proportion of individuals sampled
fine_map_file = file.path(map_dir, "example_fine.tif"), # the path to the fine resolution map
coarse_map_file = file.path(map_dir, "example_coarse.tif"), # the path to the coarse map
sample_mask_file = file.path(map_dir, "example_mask.tif"), # the path to the sample mask
partial_setup=TRUE # Don't fully import the maps yet - more to be added
)
# Add a historical map
simulation$addHistoricalMap(historical_fine_map =file.path(map_dir, "example_historical_fine.tif"),
historical_coarse_map = file.path(map_dir, "example_coarse.tif"))
# Run the actual simulation
simulation$runSimulation()
# Post-simulation
# Apply additional speciation rates
simulation$applySpeciationRates(speciation_rates = c(0.001, 0.7, 0.8))
# Output to a database
simulation$output()
# Get the species richness
simulation$getSpeciesRichness(1)- Sam Thompson (Imperial College London / National University Singapore)
- Based on ideas and code provided by James Rosindell (contact information on request).