Open source workflow in R for spatial data and analysis

[jamiecmontgomery]

What are the best workflows and packages for spatial data and analyses?

We do a lot of geospatial processing in the Ocean Health Index (see Issue #28), and we would like to streamline our workflow and to ensure all our analysis is done on open source platforms, ideally R. We currently do some spatial analyses in R using packages such as raster, sp, maps, rgdal, and more, but we still rely heavily on Python and ArcGIS because of limitations we have encountered in R (e.g., dealing with large shapefiles, etc.).

If this is a problem others encounter often enough, it would be great to develop an effort to do/discuss/improve either or all of the following:

• Is it possible to develop methods so spatial vector objects (i.e., shapefiles, etc.) aren’t read into the working memory, similar to the way that the raster package deals with raster files?
• Identify limitations to spatial analysis in R and identify workarounds to these issues
• Best approaches for integrating QGIS/Python/R

If this gains interest I'm happy to have this topic take a form that's most conducive to the community