Skip to content

Latest commit

 

History

History
18 lines (13 loc) · 1.02 KB

ENV_NOTES.md

File metadata and controls

18 lines (13 loc) · 1.02 KB

Environment Configuration Notes

This analysis is dependent upon a matrix of complicated dependencies due to the complexities of dealing with raster data. This document intends to document some of those configuration shenanigans.

AWS Notes

Instance Type: c4.large / c4.xlarge. AMI : At first just the Amazon Linux AMI, but then the Deep Learning (Ubuntu) AMI.

Package Notes

  • Don't install geopandas/fiona with conda in the tensorflow_p36 environment. It doesn't work (ImportError from Fiona, some GDAL dependency not being accounted for. Seems like pip can handle it though.
  • Installing rasterio on the Deep Learning AMI was a bit of a bear. I think because of a complex GEOS/GDAL dependency we need to build from source:
    • sudo add-apt-repository ppa:ubuntugis/ppa
    • sudo apt-get update
    • sudo apt-get install gdal-bin libgdal-dev
    • pip install rasterio --no-binary rasterio (important to build from source)
  • Seems like it doesn't matter if rasterio or geopandas is installed first.