mapme.pipelines

The codes in this repository are currently WIP. The goal of this project is to conduct large-scale analysis of globally distributed portfolios based on reproducible pipelines powered by {mapme.biodiversity}.

To reduce the production of boilerplate code when using conducting analysis with mapme.biodiversity one can instead produce a high-level YAML file indicating input/output files, options as well as which resources to fetch and indicators to calculate.

To install the package run:

remotes::install_github("mapme-initiative/mapme.pipelines")

The package exports a single function called run_config() which you should point towards a YAML file. Suppose you wanted to run the calc_treecover_area() indicator for a GeoPackage called my-polygons.gpkg, the yaml should look something like this:

input: ./my-polygons.gpkg
output: ./my-polygons-treecover.gpkg
datadir: ./data
options:
  maxcores: 4
  progress: true
  chunksize: 50000
resources:
  get_gfw_treecover:
    args:
      version: GFC-2023-v1.11
  get_gfw_lossyear:
    args: 
      version: GFC-2023-v1.11
indicators:
  calc_treecover_area:
    args: 
      min_cover: 30
      min_size: 1

Putting the content above in a file called config.yaml, all we have to do to run the configured pipeline is:

library(mapme.biodiversity)
library(mapme.pipelines)
run_config("./config.yaml")

See help(run_config) for additional details how to customize your pipeline.