The COmplex Data EXplorer is a web-application that was developed to help scientists and operators at NASA JPL perform quick explorations of high-dimensional data-sets to find patterns or relationships in the data and prepare it for in-depth follow-on analysis.
The user guide is in the docs folder of this repo, in markdown format. It covers what to do once you get the web application up and running.
There are three options for installing and running CODEX:
These two main install options are covered below.
Installation on a server is the same as for a local machine, with the additional step of opening up access to remote users and creating a DNS entry for ease-of-use. Remote-server installation is not covered here; it will be straightforward for someone who typically manages web servers.
A version of CODEX for development can be run in Docker for ease of setup. (You'll need to install Docker on your system first.)
Note that CODEX requirest 4GB of RAM to run, the default Docker RAM allocation is smaller, so you will need to adjust this in the Docker container settings. More information about how to do this can be found here
To start, run docker-compose up in the root directory of this repo. Docker will build two different images, one for the backend (using the Dockerfile in server/) and one with the web client (using the Dockerfile in client/). Note that the installation and build may take a while to complete.
The client will be available at http://localhost:8000. (port selection can be set in the docker-compose.yaml file.)
Development note:
The development build can be run by using the following command:
docker-compose -f docker-compose.dev.yml upgit clone https://github.jpl.nasa.gov/jackal/CODEX.git
Run npm install in client/ directory. (you need Node.js installed)
Set a CODEX_ROOT environment variable to point into GIT/server/
e.g., in bash something like
export CODEX_ROOT="/Users/myname/Documents/Projects/github/CODEX/server/"
Ensure you have the codex conda environment set up on your machine. Follow instructions here to do this for the first time.
Activate the conda environment, compile the client and start the server. run_codex.sh will run both the client and server in a single terminal. If you would like to start each individually, as reccomended for debugging, see the note below.
in server/ run
conda env update -f envs/environment.yml
$ conda activate codex
$ GIT/bin/run_codex.shNote: To start the client and server in different terminal sessions, follow instructions below.
Terminal 1:
$ cd GIT/client/
$ npm start
Terminal 2:
$ conda activate codex
$ cd $CODEX_ROOT
$ python codex.py