In Naylor and Jeffries (2006) we introduced a new maximum likelihood statistic, tau-squared. Although its clear this can have widespread applications, for the moment we are concentrating on developing it to fit colour-magnitude diagrams. The technique is described in that paper, though there are some improvements we have made since which are described in Naylor (2009). You can see our applications of it to measure ages and distances to young clusters in Bell et al (2014), Bell et al (2013), Littlefair et al (2010), Naylor (2009), Jeffries et al (2009), Mayne & Naylor (2008) and Jeffries et al (2007). It has been used by other groups for other clusters by Curtis et al (2013), Cargile & James (2010) and Joshi et al (2008). Finally Da Rio et al (2010) have extended the method to measure ages spreads in the Orion Nebula Cluster.

Our aim here is to provide the where-with-all for you to fit data. The software is still being developed, so you may well find you cannot just pick it up and use it on your favourite dataset. But we hope to provide you with enough that given a little thought, you can carry out fits with relatively little effort. In that spirit, therefore, we provide software that you can download, a user manual, and some worked examples. Equally importantly, though, we provide the Fortran code, so you can change the software to achieve your aims.

The software is not yet user friendly, and watch it like a hawk if you move away from the (well trodden) path of particular parameters and isochrones Rob and I have used. On the other hand, I'm happy to help, and welcome suggestions (especially in the form of revised code). Watch out for pitfalls on the hints and tips pages, and if you start using the software, drop me your e-mail address and I'll undertake to keep you informed of developments.

So, you should first go to the downloads section, install the software, and then try the examples.

When it comes to fitting your own data, I strongly advise that you first use the program iso to create some simple isochrones that you can plot over your data, to get some idea of where the parameters lie. It will also allow you to find any outlying data points which should be removed before you attempt a tau-squared fit.

Tim Naylor (timn[@]astro.ex.ac.uk)