Roy Triesscheijn’s Weblog

My programming world

Another faster version of A* (2D+3D) in C#

Posted by Roy Triesscheijn on 24th September 2011

As you might know I once wrote an A* sample for C# more than 2 years ago. The first version worked but was very slow because of a bug. The second version was faster, and a third version (made by one of the readers of this blog) was even faster.

Btw did you know that the excellent game Dysomnia shipped with (a modified) version of my C#/A* code? Ok it’s just a few lines to get people started but it’s really cool to see how people use this little start for their cool projects!

Anyway, back on topic: during my internship at Nixxes an interested co-worker (thanks Marcel!) tweaked the heuristic, removed some unnecessary checks, and added a faster way to check if a tile was processed yet.

There is not much else to say, it’s still A*, it supports 2D and 3D, it uses a MinHeap and is faster than ever! (Note: this is a pure C# solution, so it doesn’t require XNA, but it will work nicely with XNA even on WP7 and the Xbox360).

Download the latest version here

Older relevant blog posts about A*  (from older to newer)
http://roy-t.nl/index.php/2009/02/24/implementing-a-path-finding-in-c-and-xna-source-code-can-we-cut-the-corner/
http://roy-t.nl/index.php/2009/07/07/new-version-a-pathfinding-in-3d/
http://roy-t.nl/index.php/2010/06/27/more-a-improvements/

Tags: , , , , , ,
Posted in Blog, General Coding, General Gamedesign, Tips, XNA | 4 Comments »

More A* improvements

Posted by Roy Triesscheijn on 27th June 2010

As you might know I’ve written quite a lot of articles on the A* algorithm.

This week I was contacted by Roman Kazakov (fourfor[AT]hotmail.com). He had a look at my previous A* articles and came up with a number of ways to speed up my latest version even more! Unfortunately he doesn’t have a blog, so he asked me to publish his enhancements here, so that everyone can use them. So I’ll forego the rest of this introduction and let the man speak.

(Note: the following quote is the last unedited part of an e-mail of Roman Kazakov).

Hi Roy, my name is Roman Kazakov and I am Russian guy, but live in Latin America. My Spanish is perfect, but my English is “so so”, I hope you understand me. :) I am not a professional programmer. This is just my hobby from university. And as “free lancer” I work with VS 2008 C# + DevXpress + SQL Server 2008.
I would like to present you my optimization for you program AStar. Oh, by the way, my game is 2D, and this example is in 2D dimension, but this version will by work and 3D too and more faster.
Here is code of FindPathReversed:

 private static BreadCrumb3 FindPathReversed5Final(bool[,] worldBlocked, Point2D start, Point2D end)
        {
            // Here we dont use the class Poin2D. It is bad idea use Class just for two int values (X Y)
            List<BreadCrumb3> openList2 = new List<BreadCrumb3>(256);
            BreadCrumb3[,] brWorld = new BreadCrumb3[worldBlocked.GetLength(0), worldBlocked.GetLength(1)];
            BreadCrumb3 node;
            int t1, t2;
            int tmpX, tmpY;
            int cost;
            int diff;
            t1 = worldBlocked.GetLength(0);
            t2 = worldBlocked.GetLength(1);
            BreadCrumb3 current = new BreadCrumb3(start.X, start.Y);
            current.cost = 0;

            BreadCrumb3 finish = new BreadCrumb3(end.X, end.Y);
            brWorld[current.X, current.Y] = current;
            openList2.Add(current);

            while (openList2.Count > 0)
            {
                //Find best item and switch it to the 'closedList'
                current = openList2[0];
                openList2.RemoveAt(0);
                int position = 0;
                cost = openList2.Count;
                do
                {
                    int left = ((position << 1) + 1);
                    int right = left + 1;
                    int minPosition;

                    if (left < cost && openList2[left].cost < openList2[position].cost)
                    {
                        minPosition = left;
                    }
                    else
                    {
                        minPosition = position;
                    }

                    if (right < cost && openList2[right].cost < openList2[minPosition].cost)
                    {
                        minPosition = right;
                    }

                    if (minPosition != position)
                    {
                        node = openList2[position];
                        openList2[position] = openList2[minPosition];
                        openList2[minPosition] = node;
                        position = minPosition;
                    }
                    else
                    {
                        break;
                    }

                } while (true);

                current.onClosedList = true;

                //Find neighbours
                for (int i = 0; i < 8; i++)
                {
                    tmpX = current.X + surrounding2[i, 0];
                    tmpY = current.Y + surrounding2[i, 1];
                 //   tmp = current.position + surrounding[i]; //This is is a really slow!!!
                    /*if (tmp.X >= 0 && tmp.X < t1 &&
                tmp.Y >= 0 && tmp.Y < t2 &&
                !worldBlocked[tmp.X, tmp.Y])*/
                    if (tmpX >= 0 && tmpX < t1 &&
             tmpY >= 0 && tmpY < t2 &&
             !worldBlocked[tmpX, tmpY])
                    {
                        //Check if we've already examined a neighbour, if not create a new node for it.
                        if (brWorld[tmpX, tmpY] == null)
                        {
                          //  node = new BreadCrumb3(tmpX,tmpY);
                           // brWorld[tmpX, tmpY] = node;
                            brWorld[tmpX, tmpY] = node = new BreadCrumb3(tmpX, tmpY); // This is more fast!
                        }
                        else
                        {
                            node = brWorld[tmpX, tmpY];
                        }

                        //If the node is not on the 'closedList' check it's new score, keep the best
                        if (!node.onClosedList)
                        {
                            diff = 0;
                            if (current.X != node.X)
                            {
                                diff += 1;
                            }
                            if (current.Y != node.Y)
                            {
                                diff += 1;
                            }
                            cost = current.cost + diff + ((node.X - end.X) * (node.X - end.X)) + ((node.Y - end.Y) * (node.Y - end.Y));//node.position.GetDistanceSquared(end);

                            if (cost < node.cost)
                            {
                                node.cost = cost;
                                node.next = current;
                            }

                            //If the node wasn't on the openList yet, add it
                            if (!node.onOpenList)
                            {
                                //Check to see if we're done
                               // if (node.Equals(finish)) // This is slow too!!!
                                if (node.X == finish.X && node.Y == finish.Y)
                                {
                                    node.next = current;
                                    return node;
                                }
                                node.onOpenList = true;
                                openList2.Add(node);

                                int position2 = openList2.Count - 1;

                                int parentPosition = ((position2 - 1) >> 1);

                                while (position2 > 0 && openList2[parentPosition].cost > openList2[position2].cost)
                                {
                                    node = openList2[position2];
                                    openList2[position2] = openList2[parentPosition];
                                    openList2[parentPosition] = node;
                                    position2 = parentPosition;
                                    parentPosition = ((position2 - 1) >> 1);
                                }

                            }
                        }
                    }
                }
            }
            return null; //no path found
        }

Comments:
Look in my screenshots which i made it for you. For looking performance I used ANTS Performance Profiler 5.2. This is a great “stuff”!
Also you can use your benchmark. Check the file zip attached. There is all source and information for compiled.
Comments about code:
The general “thing” in programming in C# is do not use Obj Class and etc. where really dont need to use that! The best performance is just “Value type” and List<>
Array, maybe Struct.
In my optimized code I deleting class MinHeap where T : IComparable. Why?! Because is really slow, and I dont know why. Also use class Point2D just for 2(3) types X and Y( and Z), it was a bad idea. And many thing more. Looked my code and you understand which changes I made it. I hope this is help you. And please put this optimization on you wonderful web page http://roy-t.nl, as New Optimized Version A*.
When we made the games, we looking for more faster algorithm and code. This article will be help a many people as us.
We made the games of our dreams…

Roman also included a couple of pictures to show the speed gain of his code
Original code (by me)

New implementations (by Roman)

Tags: , , , , , ,
Posted in Blog, General Gamedesign, XNA | 3 Comments »

New Version A* pathfinding in 3D

Posted by Roy Triesscheijn on 7th July 2009

Intro and thanks.

Ok so let’s quickly forget the debacle around the first release of this code sample, I wan’t to delete the entire post about that one but I would’nt  have know that I made an error (until much later on) if it wasn’t for some people spotting it instantly! So I would first like to thank some of my visitors who spotted the errors:

-Oscar for spotting possible performance problems.

-Ryan for suggesting possible HashSet’s (altough I didn’t use them in the end, see below why).

I would also like to thank posters from www.tweakers.net  who helped me get this version of A* so fast, especially (on post order):

-Phyxion & Mammon for their one line version of Equals()

-Nick The Heazk for spotting that I totally forgot the Heuristic part for scoring (doh, I swear I had it in my draft)

-pedorus &  jip_86 for suggesting to use a Binary(Min)Heap

-Marcj & NotPingu for pointing out a 3D version of Pythagoras

-Soultaker for letting me think more about efficiency of containers (O(?) of add, extract and contains))

-.oisyn & pedorus for suggesting to combine a binary heap and a HashSet (altough I didn’t use it in the end, see why below)

-PrisonerOfPain for suggesting to use binary flags instead of extra HashSet’s (that’s why I didn’t use them).

Why is this version faster?

A lot was changed from the previous version, let’s start with the openList, checking if an item is in the openList costs O(n) iterations and checking which item has the lowest score costs O(n) and removing an item costs O(log(n)). So I’ve replaced the openList with a BinaryHeap for checking the lowest score and removing/adding. These operations cost O(1), O(log(n)). That’s allot faster already.

As for checking if an item is in an openList (or closedList) all I did was add a boolean flag to the BreadCrumb class (onOpenList and onClosedList).  Checking this flagg is O(1) so that really makes checks allot faster.

Also all I needed the closedList for was checking if items where already in there, so with the boolean flag I could completely remove the closedList.

Another new feature is that we immediatly return from the FindPath function when we find the finish, this also saves some operations.

I also made sure that I don’t  create to much new objects but re-use them after they where first indexed (this was  also the cause for a small ‘score’ bug) and I’ve re-aranged some ifs.

For the algorithm itself I now use DistanceSquared for the (forgotten) Heuristic part of the scoring which is allot better than ManHattan distance which is slightly biased. I’ve also changed the cost (G) to move from the current node to the next node to incorporate these squared cost  (so instead of 1,  1.4 and 1.7 I can now use 1,2,3 for these scores. I also don’t have to multiply all these numbers by 10 since 1, 2 and 3 are integers.

A final additon is the class Point3D which allows us to use only Integers instead of floats (which are slower for a cpu).

All in all this made this code about 200x faster (yes really!). But that is not completely fair since the first code was broken. If you count from after I fixed the heuristic part of the code the code is ‘only’ about 30x faster.

Benchmark results.

And before we get to the exciting part (the code!).  Let’s first show some benchmarks.

Situation:

World: 10x10x10

Start: 0,0,0

End: 9,5,8

Nodes blocked: 300

Time:

Total (100 iterations) : 134ms

Average:  1.34ms

Lowest:   < 1ms

Highest:  17ms

Note: as you can see because we did this test 100 times there is quite allot off difference between the lowest and highest time we needed to calculate this route, that’s because sometimes the cpu stops executing this program, and starts handleing an irq or doing something else, that’s why it’s important to always take a big number of benches to be representative. We can see from the average that this code is extremely fast.

Efficiency:

A* is an extremely efficient algorithm, consider we would’ve liked to brute-force this problem instead of using A*. That would’ve cost us 10^10^10 = 1,e+100 iterations. However with an (this) A* implementation we only needed 119 iterations (in which we checked 3094 nodes).

So, on to the code!

The code.

Well since I’ve divided every thing in much neater classes it’s a bit of a problem adding showing off all these as code-listings as I usually do. So I will only show the most important method here, the rest of the classes, and an example  you can download at the bottom of this article or by clicking here

        /// <summary>
        /// Method that switfly finds the best path from start to end. Doesn't reverse outcome
        /// </summary>
        /// <returns>The end breadcrump where each .next is a step back)</returns>
        private static BreadCrumb FindPathReversed(World world, Point3D start, Point3D end)
        {
            MinHeap<breadCrumb> openList = new MinHeap<breadCrumb>(256);
            BreadCrumb[, ,] brWorld = new BreadCrumb[world.Right, world.Top, world.Back];
            BreadCrumb node;
            Point3D tmp;
            int cost;
            int diff;

            BreadCrumb current = new BreadCrumb(start);
            current.cost = 0;

            BreadCrumb finish = new BreadCrumb(end);
            brWorld[current.position.X, current.position.Y, current.position.Z] = current;
            openList.Add(current);

            while (openList.Count > 0)
            {
                //Find best item and switch it to the 'closedList'
                current = openList.ExtractFirst();
                current.onClosedList = true;

                //Find neighbours
                for (int i = 0; i < surrounding.Length; i++)
                {
                    tmp = current.position + surrounding[i];
                    if (world.PositionIsFree(tmp))
                    {
                        //Check if we've already examined a neighbour, if not create a new node for it.
                        if (brWorld[tmp.X, tmp.Y, tmp.Z] == null)
                        {
                            node = new BreadCrumb(tmp);
                            brWorld[tmp.X, tmp.Y, tmp.Z] = node;
                        }
                        else
                        {
                            node = brWorld[tmp.X, tmp.Y, tmp.Z];
                        }

                        //If the node is not on the 'closedList' check it's new score, keep the best
                        if (!node.onClosedList)
                        {
                            diff = 0;
                            if (current.position.X != node.position.X)
                            {
                                diff += 1;
                            }
                            if (current.position.Y != node.position.Y)
                            {
                                diff += 1;
                            }
                            if (current.position.Z != node.position.Z)
                            {
                                diff += 1;
                            }
                            cost = current.cost + diff + node.position.GetDistanceSquared(end);

                            if (cost < node.cost)
                            {
                                node.cost = cost;
                                node.next = current;
                            }

                            //If the node wasn't on the openList yet, add it
                            if (!node.onOpenList)
                            {
                                //Check to see if we're done
                                if (node.Equals(finish))
                                {
                                    node.next = current;
                                    return node;
                                }
                                node.onOpenList = true;
                                openList.Add(node);
                            }
                        }
                    }
                }
            }
            return null; //no path found
        }

(note the class also has a normal FindPath() method that switches start and end for you).

Download.

In case you missed the download link in the middle of the text, download the entire example:  here

kick it on GameDevKicks.com

Tags: , , , , , , , , , ,
Posted in General Gamedesign, XNA | 26 Comments »