Doubt in pair sample analysis

While pre-processing any time series data or data with paired samples, how can we annotate the pool samples to see if there is any single cluster of the data points?