def search()

in LaNAS/Distributed_LaNAS/server/MCTS.py [0:0]
46 lines of code
9 McCabe index (conditional complexity)

    def search(self):
        address = ('XXX.XX.XX.XXX', 8000)
        server = Listener(address, authkey=b'nasnet')

        while len(self.search_space) > 0:
            self.dump_all_states()
            print("-"*20,"iteration:", self.ITERATION )

            #dispatch & retrieve jobs:
            self.dispatch_and_retrieve_jobs(server)
            
            #assemble the training data:
            self.populate_training_data()
            print("populate training data###", "total samples:", len(self.samples)," trained:", len(self.DISPATCHED_JOB)," task queue:", len(self.TASK_QUEUE) )
            self.print_tree()

            #training the tree
            self.train_nodes()
            print("finishing training###", "total samples:", len(self.samples)," trained:", len(self.DISPATCHED_JOB)," task queue:", len(self.TASK_QUEUE) )
            self.print_tree()
            
            #clear the data in nodes
            print("reset training data###", "total samples:", len(self.samples)," trained:", len(self.DISPATCHED_JOB)," task queue:", len(self.TASK_QUEUE) )
            self.reset_node_data()
            self.print_tree()
            
            print("populate prediction data###", "total samples:", len(self.samples)," trained:", len(self.DISPATCHED_JOB)," task queue:", len(self.TASK_QUEUE) )
            self.populate_prediction_data()
            #self.print_tree()
            
            print("predict###", "total samples:", len(self.samples)," trained:", len(self.DISPATCHED_JOB)," task queue:", len(self.TASK_QUEUE) )
            self.predict_nodes()
            self.check_leaf_bags()
            self.print_tree()
            
            for i in range(0, 50):
                #select
                target_bin   = self.select()
                sampled_arch = target_bin.sample_arch()
                sampled_arch = None
                #NOTED: the sampled arch can be None
                if sampled_arch is not None:
                #TODO: back-propogate an architecture
                #push the arch into task queue
                    if json.dumps(sampled_arch) not in self.DISPATCHED_JOB:
                        self.TASK_QUEUE.append( sampled_arch )
                        self.DISPATCHED_JOB[json.dumps(sampled_arch)] = 0
                        self.search_space.remove(sampled_arch)
                else:
                    #trail 1: pick a network from the best leaf
                    for n in self.nodes:
                        if n.is_leaf == True:
                            sampled_arch = n.sample_arch()
                            if sampled_arch is not None:
                                if json.dumps(sampled_arch) not in self.DISPATCHED_JOB:
                                    self.TASK_QUEUE.append( sampled_arch )
                                    self.DISPATCHED_JOB[json.dumps( sampled_arch )] = 0
                                    self.search_space.remove(sampled_arch)
                                    break
                            else:
                                continue
            self.print_task_queue()
            self.print_tree()
            self.ITERATION  += 1