I’m looking to move as much of my workflow as I can into julia from R/Python, and was hoping to get a sense of what the Julia GIS ecosystem looks like (especially for vector data).
Has Julia developed anything like sp in R or shapely/geopandas in Python?
(last state of the ecosystem post is a couple years old)
To my knowledge, the closest thing to shapely is LibGEOS.jl and the closest thing to fiona+rasterio is ArchGDAL.jl
I am still in the midst of designing a more user-friendly API, but if you’re looking to get started now, Andrew has started a series of posts on using ArchGDAL: GitHub - acgeospatial/Julia_Geospatial: Examples for a blog series on Geospatial Julia using ArchGDAL
Hi,
I experimented a lot with traditional GIS APIs (e.g. QGIS, GRASS, SAGA) to get them work with Julia. They cover much more than native R and Python packages and have fairly usable APIs. I didn’t managed to get it working flawlessly using the Python APIs, especially without reducing a lot of memory. What I can recommend is to use SAGA’s C API as much as possible for all kind of analyses (raster and vector).
BTW have a look at GeoStats.jl. It certainly offers some nice features for Kriging but is still less complete than what is accessible via the SAGA C API.
Can you elaborate on this? I have not heard of SAGA before, and it seems like they use Proj4 and GDAL under the hood, so I’m interested to learn more about it. Where should users go looking for documentation and examples, etc
Thank you @SimonW for sharing GeoStats.jl. Could you please elaborate on what you think would be a good addition? Maybe open a feature request? Like @yeesian I am not familiar with the SAGA project.
Saga is a C++ project. Didn’t know it has a C API. Where are the docs?
Here’s a pretty extensive python lib for/geostats GIS http://pysal.org/
Might be some ideas in there.
@datnamer thanks for sharing. Strictly speaking, PySAL is not a geostatistics library. The same could be said about many other GIS packages out there.
Sure, but it has a pretty large geostats component, was my impression.
Hi @datnamer you mean the PySAL model in this page? http://pysal.org/documentation
If there is more than that, please let me know.
I guess so. I don’t know much about geostats, so I take it that isn’t quite extensive?
Yes, geostats is a lot more 
