Applicative updating and provisional computation in functional programming

Large scale computing systems often consist of calculations which are repeated for many sets of input data. If the variance between the data sets is small, there may be portions of the computation which are not affected by the changes in the input values. The effort required for such systems can be reduced with efficient methods of recomputation. The dissertation presents an approach for efficient recomputation over function graphs. A change to an input value of a computed function initiates a process of retraction along the data paths dependent on that input. The modified value is then released into the function graph, and the retracted portions of the graph are recomputed. The mechanism for retraction maintains functionality of output values and provides an efficient method of recomputation. The mechanism is extended to include a new class of operators which provide provisional computation by initiating retractions wholly within a function graph. This extension addresses the problem of feedback cycles in a computation and provides an internal mechanism for detecting and breaking deadlock encountered during the execution of such cycles. A retraction mechanism which allows efficient recomputation increases ·the use fullness of functional programming systems. The effort required to carry out the retraction is proportional to the number of arcs in the function . graph. Allowing provisional computations expands the class of graphs which will produce results by allowing feedback cycles to be expressed naturally.