This sample db demonstrates identification of clusters in value series as per given maximum allowed spacing between adjacent elements of a cluster. For this, sorted values are subjected to sliding comparison via unequal self outer join.

    Original data is displayed in the subform at left. Based upon user selection of SeriesID and maximum allowed spacing between adjacent elements of a cluster, the values qualifying as members of various clusters are displayed in the subform at right. Start value for each cluster is also displayed in a separate column.

    Simultaneously, matching value elements in in the subform at left get highlighted in distinct color.