python (numpy) implementation of Powell's optimisation algorithm

As the title says, we are looking for an implementation of Powell's algorithm for non-linear optimisation without derivatives using python / numPy. A maths background is recommended. The code should be very readable and well-commented, all variables explained. As a line search algo, please let the user choose between Golden section and Quadratic Fit. Maybe a flexible linesearch object is a good idea. The user provides a module uf with a function Q: R^n->R that is to be optimised. _ __ ____ ______ ______________________________ We received some question WHICH algorith we mean. It is about derivative-free optimisation of a function f:R^n -> R. Every STEP consists of n (=nr of dimensions) line searches. For the first step, let every direction_i be the i-th base direction. After this, compare the start vector and the vector after the first step. The difference is called P_0. Before the second step, set every direction direction_i to direction_[i-1], and set direction_1 to P_0. And so on. After a certain number of steps (user set, default to 2*n), the directions may become linear dependent. therefore, the direction set is than resetted to the base directions. The user can decide on: reset step nr (defaults to 2*n). Line search algo (Golden section or QF). Termination criterion for algo. Termination criterion for every line search. ## Deliverables 1) Complete and fully-functional working program(s) in executable form as well as complete source code of all work done. 2) Deliverables must be in ready-to-run condition, as follows (depending on the nature of the deliverables): a) For web sites or other server-side deliverables intended to only ever exist in one place in the Buyer's environment--Deliverables must be installed by the Seller in ready-to-run condition in the Buyer's environment. b) For all others including desktop software or software the buyer intends to distribute: A software installation package that will install the software in ready-to-run condition on the platform(s) specified in this bid request. 3) All deliverables will be considered "work made for hire" under U.S. Copyright law. Buyer will receive exclusive and complete copyrights to all work purchased. (No GPL, GNU, 3rd party components, etc. unless all copyright ramifications are explained AND AGREED TO by the buyer on the site per the coder's Seller Legal Agreement). * * *This broadcast message was sent to all bidders on Wednesday Jun 11, 2008 3:54:36 AM: Dear collegues, I added a rough algo description to indicate clearer what I mean. i assume the thing is quite straight forward to do in numpy. Regards ## Platform plattform-indipendent. python with scipy/numpy.