Dynamic data redistributions in data-parallel languages like HPF
address the demands posed by advanced applications with different
computational kernels or dynamically varying processor workloads
and are a major means for improving the performance.  On
the other hand, redistributions can be very expensive and significantly
degrade a program's performance. In this article, we present a powerful
and flexible approach for avoiding unnecessary remappings by
eliminating partially dead and partially redundant
distribution changes. Basically, this approach evolves from extending
and combining two algorithms for these optimizations for sequential
programs. In contrast to the sequential setting, the data-parallel
setting leads to a hierarchy of algorithms of varying power and
efficiency. The flexibility resulting from this makes it easy to trade
efficiency against power according to a user's requirements.  The
approach is demonstrated by several illustrating examples.