[Skip to Content]



Clusterix is simple, yet sophisticated tool designed to separate open cluster's stars from those, that belong to stellar background. In opposition to the classical approach, non-parametric methodology is used in this study (see 'about' link for further information). To retrieve results, simply fill in the form below or find required cluster using "Lists of open clusters" link on the left side of this page.

Cluster search help

File upload

If you are interested in data which are not reachable via offered databases, you may inspect your own data set. All you have to do is create it in accordance with specification which is described in this sample file and submit your request via following form. Data set has to contain at least four stars, maximum file size is set to 20 MB.

Upload form

In a nutshell

Astrophysics offers a plethora of problems whose solutions may be retrieved almost exclusively by using the computational power. The reasons are obvious: thousands of astronomical surveys performed during last several decades produced enormous amount of data (more than 10^10 stellar object are catalogued so far) whose manual processing is beyond the man's possibilities. Also, mathematical models needed to deal with emerging challenges are often very complex and usage of their numerical representation is inevitable.

empirical probability density function visualisation

Determination of open cluster membership probability is an example of such a problem. Using the statistical methodologies, stars surrounding central coordinates of an alleged cluster may be investigated in order to determine their origin. Such a cluster-field division is requisite as it allows us to form conclusions about separated populations - knowing the cluster members, almost all fundamental properties of examined stellar association (such as age, distance, etc.) may be retrieved via standard astronomical methods.

Basic idea of the non-parametric method for the cluster-field separation in the 2-dimensional proper motion space is the empirical determination of the cluster and field stars distributions without any assumption about their shape. Cabrera-Cano & Alfaro (1990) used the kernel estimation technique (with a circular Gaussian kernel function) to derive the data distributions. Different implementations of this approach were further described by Chen et al (1998) or Galadí-Enríquez et al (1998). [ L. Jílková, 2013 ]