Docker-ISIS

Container images for geo-planetary data analysis.

This repository defines a Docker containers with USGS/ISIS and NASA/ASP toolkits ready for use, as well as a GIS environment based on Python and command-line tools for either Earth or planetary geo-spatial data processing.

The goal is to remove the time-consuming and error-prone steps on setting up those data analysis/processing software. Besides optimizing the time spent on setup, a standard set of images provide reproducible scenario for more reliable results.

All the (three) images extend on Jupyter Docker Stacks images, meaning they all provide a Jupyter Lab interface for high level access to Python and CLI.

The (Docker) image recipes we provide here are:

• gispy: Jupyter-Lab for Geographical (GIS) data analysis with Python
• isis: Jupyter-Lab server with USGS/ISIS installed
• isis-asp: Extension of isis with AMES Stereo Pipeline installed

Ready-for-use images can be downloaded from GMAP repository in DockerHub:

• gmap/jupyter-gispy: latest build of gispy.dockerfile
• gmap/jupyter-isis: latest build of isis.dockerfile
• gmap/jupyter-isis-asp: latest build of isis-asp.dockerfile

Images tree:

|---------|     ---------
| Jupyter | --> | gispy |
| Docker  |     ---------
| Stacks  |     ---------     ------------
| images  | --> | isis  | --> | isis-asp |
|---------|     ---------     ------------

If you want to build your own images, go to dockerfiles/README.md.

How to use it

Since these images are based on Jupyter images, running them works exactly the same:

docker run -it -p 8888:8888 -e JUPYTER_TOKEN=pass gmap/jupyter-isis

Open a browser window at http://localhost:8888 and type pass in the password/token field for the authentication. Notice that we used the argument -e JUPYTER_TOKEN=pass to set the password/token. If you don't use such argument, you should get the token value from the (Jupyter) service log from docker run command.

Since most of ISIS tools demand the use of ISIS-Data, you need to provide it by sharing a local directory with ISISDATA (see ISIS-Data below).

Per default, the container images (ie, jupyter-isis, jupyter-isis-asp) define the ISISDATA variable to:

ISISDATA=/mnt/isis/data

Analogously, ISISTESTDATA (in case you use) is defined to /mnt/isis/testdata.

You will then bind your local directory to that of ISISDATA inside the image though docker-run -v option:

MY_ISIS_DATA="/path/to/my/isis/data/in/your/system"
docker run -it -p 8888:8888 -v $MY_ISIS_DATA:/isis/data gmap/jupyter-isis

You can share as many directories as you want (between your computer and the container) through (multiple) -v arguments. For example, if you want to analyse data there are stored in two different diretories in your computer -- say, /path/to/raster and /path/to/vector --, you can share them with the container like so:

docker run -it -p 8888:8888 \
    -v /path/to/raster:/home/jovyan/raster \
    -v /path/to/vector:/home/jovyan/vector \
    gmap/jupyter-gispy

The "raster" and "vector" (local) directories will be mounted in container's /home/jovyan/raster and /home/jovyan/vector, resp.

ISIS-Data

If using ISIS tools, you'll probably need "ISISDATA" (the space-planetary missions and instruments support data). ISIS-Data is NOT included in the container(s).

For instructions on obtaining ISIS-Data, see the official docs at:

• https://github.com/DOI-USGS/ISIS3/blob/dev/README.md#The-ISIS-Data-Area

In the next section you'll learn how to use it.