Earlier
Assumptions made about artifacts earlier than those in the data can greatly influence the outcome of any seriation. Feature values for earlier artifacts are essentially blank and like blanks it is not always obvious how to interpret them. We could assume that earlier feature values are simply unknown, or we could assume that a feature is absent for earlier artifacts or we could assume that the feature value is zero. Whatever convention we are using, we need to let OptiPath know by setting the feature parameter Earlier in the Features table.
Earlier values can be interpreted three different ways: as unknowns, as absences or as zeroes. When choosing a value for Earlier, the options are Unknown, Absent and Zero.
If the feature parameter Earlier is set to Unknown, feature values for earlier artifacts (earlier than the earliest artifact in your data) are assumed to be unknown and the feature will have no effect in the computation of distance between an artifact and those that came before the artifacts in the data set. This means this feature's contribution to the distance, including any transition penalty, is zero.
If the feature parameter Earlier is set to Absent, feature values for earlier artifacts are treated like absences and will affect the computation of transition penalties but otherwise they are treated like unknown values and have no effect in computing distance between artifacts.
If the feature parameter Earlier is set to Zero, feature values for earlier artifacts will be interpreted according to the setting for the feature parameter Zeroes. You can indicate that earlier artifacts should be assumed to have both a value of zero and an absence by telling OptiPath to interpret earlier artifacts as having a value of zero and to interpret zeroes as both a value and an absence (see Zeroes).
Each setting for Earlier can lead to different results in seriation. For more information see Setting the Earlier, Later, Blanks, Zeroes and Transition Parameters.