Dear all, for a project, I need a data structure to represent finite distribution, that is a set of elements, each of which having a mass in the set, and two functions:
- total-mass : the sum of the masses of each element - draw : return an element with a probability proportional to its mass I currently use the simple idea : a Persistent Hash Map from elt to mass: - the total-mass is the reduce of (+ . second) - drawing is choosing a random-number between 0 and the total-mass and looping through the seq, to look for the corresponding element. Both operation are linear in the number of elements. This is a problem, because these sets tend to be quite big in my project. I think that if I used a HashTrie with total-mass cached in each node, total-mass would be in constant time and draw in log time. So here is my question : does anyone has, for example for the "clojure in clojure" project, a sketch of an implementation of PersistentHashMap in clojure itself? (I would be happy to contribute, if it needs more work). Or has anyone a better idea that mine to have drawable sets? Thanks for your help, Best regards, Nicolas. -- You received this message because you are subscribed to the Google Groups "Clojure" group. To post to this group, send email to clojure@googlegroups.com Note that posts from new members are moderated - please be patient with your first post. To unsubscribe from this group, send email to clojure+unsubscr...@googlegroups.com For more options, visit this group at http://groups.google.com/group/clojure?hl=en
