Overview

The MATSS package is intended to support Macroecological Analysis of Time Series Structure. We provide functions to:

  • gather ecological time series datasets
  • perform basic processing and summaries of those datasets
  • build an analytical pipeline to conduct macroecological analyses on those datasets
  • create template reports for collating results and produce syntheses

Contributing

For more information about contributing code, datasets, or analyses, please check out the Contributing Guide.

Installation

You can install MATSS from github with:

# install.packages("remotes")
remotes::install_github("weecology/MATSS", build_opts = c("--no-resave-data", "--no-manual"))

This package relies on the development version of the rdataretriever package to install datasets. Installation of this package takes a few extra steps because it runs a Python package behind the scenes. Follow the installation instructions on the rdataretriever README.

Datasets

MATSS pulls data from a variety of sources, including:

  • 10 individual datasets that we’ve added,
  • the North American Breeding Bird Survey database (spanning 2587 separate datasets),
  • the Global Population Dynamics Database (spanning 120 separate datasets),
  • and the BioTime database (spanning 361 separate datasets).

Combined, there are 84052 individual time series across all of these data sources.

Getting Started

To get started with the data or analysis templates, we recommend you take a look at our Getting Started vignette for more details about how to interface with the datasets, use Drake to create workflows, and create research compendia.

If you have the MATSS package installed, you can also view the vignette from within R:

vignette("MATSS")

Example Use Cases

Here are some examples of using MATSS to create research compendia:

  • MATSS-LDATS applies the LDATS package to investigate changepoints in community dynamics across the datasets in MATSS
  • MATSS-Forecasting investigates which properties are associated with the predictability of population time series across the datasets in MATSS

Acknowledgments

We thank Erica Christensen and Joan Meiners for their contributions and input on early prototypes of this project. This project would not be possible without the support of Henry Senyondo and the retriever team. Finally, we thank Will Landau and the drake team for their input and responsiveness to feedback.

Package development is supported through various funding sources: including the Gordon and Betty Moore Foundation’s Data-Driven Discovery Initiative, Grant GBMF4563 to E. P. White (supporting also the time of J. Simonis and H. Ye), the National Science Foundation, Grant DEB-1622425 to S. K. M. Ernest, and a National Science Foundation Graduate Research Fellowship (No. DGE-1315138 and DGE-1842473) to R. Diaz.