Week 1

This is my first week visiting the University of North Texas, and I'm actually very excited. I finally get to meet my mentor Dr Mehta, and get the privilege to work under her and her team on their project. I was introduced to a couple of other people in other programs that will be staying here at the university. And will in fact be working with some undergraduates from another program called SUPER.

Even though this is the first time I am Personally meeting Dr Mehta I am already familiar with the project they are working on. Dr Mehta and I had been discussing the project the month before I arrived here so I am able to just jump in the project with some sort of understanding of what we are working on.

For this week it is simple I just have to play the game and document my strategies. Also since the game is in the beta stage I will be looking for any bugs, and improvements that we could make to the game to help improve the player experience. The team as well has to work on getting interactive tutorials up and running as the current ones do very little to provide the knowledge necessary.

The game is pretty addicting but it gets really tough, and nearly impossible on some levels. I am documenting my strategies on powerpoint with screenshots to give the best representation of the strategies I have used on different levels.

Week 2

This week I will be using a depth first approach to solve the graphs in the game. This approach is used from a graph placement algorithm that I have read earlier this summer and the algorithm was pretty help full in solving the graphs in less moves, and I was able to improve a couple of the graphs I had played previously.

The second part of my work was finding bugs, and generally helping with improvements that can be made. There were a couple of bugs in the game that we were able to find, Some bug that crashed the game, and retrieve incorrect data from saved games. There was also a lot of bugs in the multi-move tool, this tool allowed players to move multiple nodes that are selected at once. The problem was by using this tool players could be able to move a node onto of another and are even able to place nodes in forbidden areas.

Week 3

This week I have been given the task of solving the scalability issues with out game. This task is broken down into two goals, finding a way to represent the harder levels that will not intimidate the players, and to Provide larger graphs that the players will be able to play with out being too intimidated (current graphs are approximately 60 nodes the larger graphs will be about 200 nodes).

This is a big challenge in the research project, and I have no idea how to even come up with a solution for this problem. To better prepare myself I have been reading over a lot of papers on placement algorithims trying to get some sort idea or technique that could used to help players solve graphs. I have also been reading a lot of puzzle design papers/resources to help with representing our harder levels so players will not get as fustrated or discouraged when approaching them.

As it stands right now nobody on our team is even able to beat the most difficult levels in the game (and that is pretty bad). But hopefully I can come up with a solution for this next week.

Week 4

This week I will be looking at clustering as a solution to the scalability problem. I will be reading alot of papers that discusses clustering algorithms for graphs. Hopefully one of the algorithims examined could be used to help us draw our graphs in a better way for the more difficult levels.

Week 5

This week I have been given a new task for our project, after examining clustering as a solution for our sclabilty problem it was clear that it was not going to work. The scalability problem was a great challenge and I struggled quite a bit with.

So my new task is creating a sub game for the one we already have. With the game our players are playing right now we assume that every cell in the grid is able to provide every operation needed for the node that it is given. My task is to create a game where specific cells provide specfic operations so only certain nodes could be placed there. For this game we would like to see how well players can play given an heterogeneous architecture. Players right now are given a homogeneous architecture.

Week 6

I have completed the prototype for the new Game type that models a heterogeneous CGRA. This model seperates operations that needs multiplication from the rest of the operations, since multiplications is more expenisve I have submitted my case studies to Dr Mehta. I will be prototyping another architecture new model as well. Hopefully this one will be finish this prototype in time. I have another task to complete by next week as well, which involves examining the output of an algorithim that maps graphs to a CGRA.

Week 7

I am finally done with the new protoype, but I still have quite a few case studies to perfrom on it as well. For now I will be examining the simmulated annealing algorithim’s output for improvements. I will also be showing this new strategy I have been using on the game to Dr Mehta that involves clustering nodes.

Week 8

This is my final week here at UNT. I have completed most of my tasks. Right now I have to finish my reserch paper, and compile all the work I have done into one folder to be sent to Professor Mehta along with a readme file.