Distribution of the calculation of a function over the nodes. Derive from it and reimplement : - treate : will be called by the MASTER each time a point is computed. this function will typically store it, and possibly affect the list of points waiting to be computed, as returned by next... NB : "None" result is ignored. - next : next point to compute. If it returns None (which is not the same as finished()) the computation is ignored. - finished() : Whether the calculation is finished - the_function : function to be computed, with the argument given by next() In the module DistributionOnNodeTest is a test example.


A special of distribution, when one has a given stack of points to compute Reimplement just the_function