Other important and convenient agglomeration functions over time are weighted
where is a weighting function with almost everywhere. This can easily be generalized to semi-norms in which may hold on a subset of [0, T] with nonzero measure, but else. Weighting functions with this property can also be introduced in (5.1). When interpreting e as a sampling function, this means that some points (in time) are left unsampled. The corresponding agglomeration functions include
i. e. only the effect at the end of the time interval considered is accounted for.
Since there are too many different time agglomeration functions which are reasonable in different circumstances, no attempt has been made to provide the user of the software package with a number of templates. If agglomerations different from the one presently implemented are preferred, the source code of the optimization module has to be changed accordingly.