# Archive of the 3 Phase Oil Data

The 3 phase oil data was data used to demonstrate models like the GTM. The page where it was hosted http://www.ncrg.aston.ac.uk/GTM/3PhaseData.html seems to have disappeared, so I’m putting links to the parts I have here. This information comes to you courtesy of the Wayback machine (http://www.archive.org).

Here is the original text from the page (the links won’t work apart from those to the data set).:

## 3 Phase Data

### Description

This is synthetic data modelling non-intrusive measurements on a pipe-line transporting a mixture of oil, water and gas. The flow in the pipe takes one out of three possible configurations: horizontally stratified, nested annular or homogeneous mixture flow. The data lives in a 12-dimensional measurement space, but for each configuration, there is only two degrees of freedom: the fraction of water and the fraction of oil. (The fraction of gas is redundant, since the three fractions must sum to one.) Hence, the data lives on a number of ‘sheets’ which locally are approximately 2-dimensional.

At the moment, there is no further information online. Even the old Aston paper repository has disappeared so I can’t any longer link to the papers that used the data from there.

### Details

The data is available either as a tar-file containing compressed ASCII files, or as a compressed MATLAB (R) workspace. The files/variables contain:

• Data*, 1000 measurements, 1000-by-12
• Data*Frctns, the corresponding fractions of water and oil (in that order), 1000-by-2
• Data*Lbls, the corresponding configuration labels, given in a 1-of-3 coding scheme, where
• [1 0 0] == Homogeneous configuration
• [0 1 0] == Annular configuration
• [0 0 1] == Stratified configuration

1000-by-3

’*’ above is replaced by ‘Trn’, ‘Vdn’, and ‘Tst’, which are meant to correspond to training, validation and test data; the three file sets all contain 1000 samples. The fractions and configurations are picked at random from corresponding uniform distributions. gzipped tar-file\ gzipped MATLAB workspace

This page is maintained by Markus Svensén (svensen@cns.mpg.de) Copyright © Markus Svensén 1996