README.md

In mtalluto/mbmtools: Multifaceted Biodiversity Modelling

MBM: Multifaceted Biodiversity Modelling

R Package for simultaneously modelling multiple facets of biodiversity (α- and β-diversity, taxonomic, functional, and phylogenetic).

Installing MBM

Before installing MBM, it is necessary to have installations of R, Python and GPy. Please install these packages and verify they are working before installing MBM. New users and those unfamiliar with Python should see the "Recommended Installation Procedure" below.

Additionally, some functionality (namely, svgp for fitting large models) requires the climin package in Python. Thus, it is highly recommended to install this package when instally GPy. Note that climin pre-0.1 is the only supported version. Thus it is necessary to follow these instructions to install climin from Github.

Within R, MBM can be installed directly from github using the devtools package.

# uncomment the line below if devtools is not yet installed
# install.packages('devtools')

library('devtools')
install_github('mtalluto/mbm')

library('mbm')
?mbm

Alternatively, if you would like additional features (principally, vignettes), you can clone or download the project and build and install it (from the command line):

git clone https://github.com/mtalluto/mbm.git
Rscript -e "devtools::build('mbm')"
R CMD INSTALL mbm_i.j.k.tar.gz

Replacing the i.j.k with the version number you have downloaded; this file will have been created by devtools in the second step.

Recommended Installation Procedure

1. Install R and Rstudio. I recommend using Rstudio for interacting with R; any time these instructions refer to doing something in R, you can do the step in Rstudio instead.
2. Install the Anaconda distribution of Python 3.7.
3. Install climin. Note that installing the default version with pip will not work, you must install the github version.
4. Install GPy; the easiest way is with pip from the command line: run the command pip install GPy. If you have errors, you can check the linked page for installation instructions.
5. Install devtools and mbm within R; follow the instructions above under Installing MBM.
6. Test that the python system is working; in R, load mbm with library("mbm") then run check_python(). If all works, you can proceed with the examples.

Macos Installation Notes

The following procedure has been tested on a clean mac running macos 10.13 using the system Python (version 2.7). From the command line:

sudo easy_install pip
sudo pip install -U --ignore-installed numpy scipy
sudo pip install GPy tornado climin

From there, assuming everything proceeds without errors, you can install and test MBM.

mtalluto/mbmtools documentation built on Aug. 13, 2019, 9:44 a.m.