Mathematica has a lot methods and options for cluster analysis (and I learned them not bad =)
Now I need to find clusterization of some daily weather data,
and the number of days in any cluster should be more than some threshold (e.g. 30).
So if some cluster has less members, it has to be merged with other.
Number of clusters itself is not defined initially, so method should not be "KMeans" or "KMedoids".
I still don’t understand what options (for e.g. "DBSCAN", "Agglomerate") can be used in this case.
