geospade

Description

geospade stands for geospatial abstract definition environment. It serves as place to define classes and properties of raster and vector geometries and their operations alike. On a higher level, abstract definitions should be embedded in a geospatial context to support interaction with other Python packages, e.g. gdal, geopandas or xarray. In comparison to these Python packages, geospade never touches or writes any geospatial data on disk. It is a toolkit for geospatial entities (e.g., points, polygons, ...) and their relations (e.g., intersection, within, ...) in a spatial reference system (e.g., reprojection, mosaics, ...).

In geospade a geospatial context is given by the spatial reference system class crs.SpatialRef, which allows to convert between different spatial reference definitions in Python (e.g., osr, cartopy.crs, ...) and offers well-known spatial reference system representations, i.e. WKT, PROJ4, and EPSG. It aims to solve discrepancies between these representations and lower-level package versions, e.g. gdal.

An abstract, geospatial definition of a raster is implemented in RasterGeometry. It is constructed by providing a pixel extent, i.e. the number of rows and columns, the 6 affine geotransformation parameters and a spatial reference system. With this knowledge, one can use a RasterGeometry instance to do many operations, e.g. intersect it with geometries, transform between pixel and spatial reference system coordinates, resize it, or interact with other raster geometries.

Often, geospatial image data is available in tiled or gridded format due to storage/memory limits. To preserve the spatial relationship for each image, MosaicGeometry can help to apply geospatial operations across image/tile boundaries. It represents a simple collection of Tile/RasterGeometry instances, where each Tile describes the spatial properties of an image. With this setup, tile relations and neighbourhoods can be derived.

Installation

The package can be either installed via pip or if you solely want to work with geospade or contribute, we recommend to install it as a conda environment. If you work already with your own environment, please have look at conda_environment.yml and install/adapt missing packages.

pip

To install geospade via pip in your own environment, use:

pip install geospade

conda

The packages also comes along with a conda environment conda_environment.yml. This is especially recommended if you want to contribute to the project. The following script will install miniconda and setup the environment on a UNIX like system. Miniconda will be installed into $HOME/miniconda.

wget http://repo.continuum.io/miniconda/Miniconda3-latest-Linux-x86_64.sh -O miniconda.sh
bash miniconda.sh -b -p $HOME/miniconda
export PATH="$HOME/miniconda/bin:$PATH"
conda env create -f conda_env.yml
source activate geospade

This script adds $HOME/miniconda/bin temporarily to the PATH to do this permanently add export PATH="$HOME/miniconda/bin:$PATH" to your .bashrc or .zshrc.

For Windows, use the following setup:

• Download the latest miniconda 3 installer for Windows
• Click on .exe file and complete the installation.
• Add the folder condabin folder to your environment variable PATH. You can find the condabin folder usually under: C:\Users\username\AppData\Local\Continuum\miniconda3\condabin
• Finally, you can set up the conda environment via:

  conda env create -f conda_environment.yml
  source activate geospade
  

After that you should be able to run

python setup.py test

to run the test suite or

python setup.py install

to install geospade.

Contribution

We are happy if you want to contribute. Please raise an issue explaining what is missing or if you find a bug. We will also gladly accept pull requests against our master branch for new features or bug fixes. If you want to contribute please follow these steps:

• Fork the geospade repository to your account
• Clone the geospade repository
• Make a new feature branch from the geospade master branch
• Add your feature
• Please include tests for your contributions in one of the test directories. We use py.test so a simple function called test_my_feature is enough
• Submit a pull request to our master branch