This repository contains code to reproduce all figures from the manuscript:

Michael Schmuker, Viktor Bahr, Ramón Huerta (2016): Exploiting plume structure to decode gas source distance using metal-oxide gas sensors. Sensors and Actuators B: Chemical 235:636-646 (2016). doi:10.1016/j.snb.2016.05.098

The accepted manuscript is available for free on ArXiv under arxiv:1602.01815.

1. Download the data that A. Vergara and colleagues have collected and published in 2013 from the UCI Machine learning Repository).

2. Extract the archive into the same top-level directory that you cloned this repository in. For example, if you clone this repo into /home/user/plume_distance/, you should extract the data into the same directory. That directory should afterwards contain at least the two entries exploiting_plume_structure (i.e., this repository), and WTD_upload (the dataset).

3. Go to the ipnotebooks directory, fire up an ipython/jupyter notebook session, and you should be good to go.

Make sure you have all dependencies installed:

• jupyter (to open the notebooks in the first place)
• numpy
• scipy
• matplot
• tables
• sklearn
• mdp
• pandas

We recommend the Anaconda Python distribution because it made scientific python a breeze on every computer and platform we were working on so far.

If you run into issues please use the issue tracker - we'll do our best to respond as quickly as possible.