Quick Start

Installation from source

  1. First you need to download and set up Anaconda or Miniconda on your computer.

Note that eViz has been tested with Python version >= 3.8

  1. Get the source code (use http protocol for read-only access, ssh otherwise):

git clone https://git.smce.nasa.gov/astg/visualization/eviz.git
git clone git@git.smce.nasa.gov:astg/visualization/eviz.git

  1. cd into the code repo:

cd eviz

  1. Create the Python environment:

conda env create -f environment.yaml

Enter y when prompted. This will download all the required packages needed to run the eViz tools and install them in a separate environment called viz. This may take a minute or two, so please be patient.

  1. Once the installation has finished building, activate the installed environment by running:

conda activate viz

Sample data

There are two eViz tools that can be from the command line interface (CLI): iViz and autoViz. The iViz tool is a Jupyter notebook-based dashboard with interactive widgets whereas autoViz is a non-interactive CLI-driven map-generating tool. In either mode, we need data. On DISCOVER we provide some data representative of the data sources supported by the eViz tools. The data is located here:

/discover/nobackup/projects/jh_tutorials/eviz/sample_data

Therein you will find datasets collected from various data sources that are used to produce the visualizations described in this guide.

You can also get the sample data from our data portal:

https://portal.nccs.nasa.gov/datashare/astg/eviz/sample_data/

JupyterHub

On Discover, eViz is installed on the JupyterHub system. To access it you must log in to:

https://jh-discover.nccs.nasa.gov/

using your LDAP credentials.

Once you get a session, open up a notebook and make sure you select the Python viz kernel. Then type the following:

from eviz.lib.iviz_base import Iviz
iviz = Iviz()
iviz.make_notebook(s='ccm')

If all goes well, you should see the iViz app running on your notebook with “ccm” data loaded and ready to browse.

Web-based plots

This is the code that we use to host visualization on a web platform. We encourage developers to try it out and experiment with setting up your own web platform.

To share your visualizations on a website, EViz offers a tool to generate plots accessible via a web browser. This functionality utilizes the streamlit package, which should already be included in the viz environment.

To use the web interface run the following command:

streamlit run sviz/sviz.py

This command will launch a web-based interface to run autoViz and display the static plots on your local host.

For additional information please look at the streamlit documentation (https://streamlit.io/).