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
For more information about contributing code, datasets, or analyses, please check out the Contributing Guide.
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.
MATSS
pulls data from a variety of sources, including:
- 10 individual datasets that we’ve added,
- the North American Breeding Bird Survey database (spanning 2589 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.
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")
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 inMATSS
- MATSS-Forecasting
investigates which properties are associated with the predictability
of population time series across the datasets in
MATSS
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.