ropensci / chopin

Computation of Spatial Data by Hierarchical and Objective Partitioning of Inputs for Parallel Processing
https://niehs.github.io/chopin/
Other
12 stars 2 forks source link

Review DeGAUSS containerized platforms #23

Closed sigmafelix closed 10 months ago

sigmafelix commented 1 year ago

TODO

Points of consideration

sigmafelix commented 12 months ago

Looking at the codes from degauss-org, I found that the container included very small pre-calculated datasets (e.g., for isochrones --polygons representing the same travel time) or pointed preprocessed datasets at S3 storage (e.g., NDVI). Each container serves a very specialized function and not other than that, where users only need to consider what they want to calculate. For our use cases, this design would be easily adopted as we mainly consider vector-raster overlay in this package. For example, a container will accept two to three parameters (of course, in terminal):

apptainer run _image.sif_ satellite_imagery.tif input_locations.gpkg 30000

where each argument is a satellite imagery, a data of input locations, and buffer radius, respectively.

sigmafelix commented 10 months ago

This platform will be referred to as an existing solution for local / study-specific geocomputation needs.