This is a weekly updated journal of what I have done with the DMP program at Texas A & M.
This week was all about setting things up. I met my student mentors Chids and Olga. My faculty advisor is Dr. Nancy Amato, but everyone calls her Nancy. There are two other undergraduate students in our group: Saransh and Harsh. Our mentors and advisor are helping us set up the interface for our graph generator, which will allow the user to create and manipulate graphs. Once everything is set up, the undergrad students (Saransh, Harsh, and myself) will divide and conquer for the implementation.
As part of the DMP program, I attend seminars, talks, and GRE sessions. The GRE sessions are to help undergrad students prepare for the test required of students trying to get into graduate school.
Week 2: June 12 - June 16
Our group members are coming and going still. Harsh came back on Monday, and we all met to go over what needs to be done while Olga is gone. I set up the task list for the undergraduates to "check out" code that we are working on so that only one person works on it at a time. Our research plans were completed and turned in. We attended the thesis defense of a Ph.D. student in the Parasol lab and had a nice group outing to celebrate. It was really good to see people from the lab outside of the Bright Building. The GRE class finished early this week because the topic was on math--something the computer science and engineering kids can handle :). I also went to a talk on "Financing Graduate School" and a presentation by a professor in the Computer Science Department. Between those workshops and brown bag lunches I created a documentation template for our group. Now, each person can just plug in the information regarding the function they have been working on without having to set up any formatting.
Week 3: June 19 - June 23
I went through the files BaseGraph.h, Graph.h, and GraphAlgo.h. A lot of the functions we use come from these files, and as I read through them I created a notes copy for later reference. Our group has been working on finding a balance between too much and too little information for the generic interface. We are at the point now where we can start testing and debugging the properties we have been implementing. The GRE session dealt with tips on some verbal sections of the test. Our brown bag on Thursday was wonderful. The speaker was there to give us a seminar on Technical Writing, and we all learned a lot from her.
Week 4: June 26 - June 30
I got the property and compositional property classes for the number of vertices and edges working. This week I have mostly been fixing compilation errors. Templates are new to me, so the boys and I are learned what to change so that inheritance works. Also, with the addition of the "properties" class (to work on all properties at once) and changing everything to the stapl namespace, the earlier code needs revision so that they all play nice together.
The Brown Bag Lunches this week were great. Wednesday we were privilegded to meet Dr. Stroustrup (designer of C++). He was a great presenter and answered lots of questions. Thrusday we had a Grad Student Panel, which allowed all of the REU and DMP students to get honest answers about the Graduate Program at TAMU. They answered questions from how they get funding to the pros and cons of B.S. vs M.S. vs Ph.D. It certainly gave me a lot to pray about. I still have time to think about it, but will probably start looking into graduate programs now.
Week 5: July 3 - July 7
This week I worked on max, min, and avg in degree properties. We were able to fix the compose functions to consider a Directed Predecessor Graph (which is the type of graph I am focusing on right now). I was introduced to this system's debugger (gdb) and starting learning how to use it. That helped me find problems that the compiler couldn't warn me about. I created and committed a test file for the group to see that the properites are finding the correct values. The compositional property versions of these in degree properties are 50% done. They return the correct values for compose no join and join at vertex.
Week 6: July 10 - July 14
This week I worked on compositional calculates for the intra merging cases and proving their accuracy. I also cleaned up the overall property classes so that the properties are listed by the type of graph they are used for. That way, it is easier to only include the properties we currently want to test. Briefly, I looked for cases of an intra join function that currently outputs error messages. That way Saransh could fix them before he left. Saransh started his summer intern work here at Texas A&M sooner than the rest of us, so naturally his ten weeks are over sooner too. Before he left, we had some code review sessions. That way everyone in the group would have a time to ask him questions before he left.
Week 7: July 17 - July 21
This week was quiet but productive. Many members of the Parasol lab were gone at a conference. I revised my research abstract and the progress report for the DMP. I also worked on a new property for our graph. I spent two days working on and testing an algorithm for it before I was reminded that there was already a function that did the computations I needed. After that property was working properly and was fully documented, I started looking for bugs in Saransh's code. I found and fixed the joining errors in intra join at edge case. On Thursday, when everyone was back from the conference, our group had a meeting with another graduate student in the department, Nathan, who gave us some direction for cleaning up the current implementation of our graph properties. My last task for the week was to read up on type traits and boost's mpl.
Week 8: July 24 - July 28
The first part of the week, I worked on moving our code over to a better implementation suggested to us by Nathan. I also organized notes on the new libraries that were revelent to what we needed for the group. After that, the rest of the week went into planning and making our poster. It is almost ready for printing for the REU Poster Presentation on Friday.
Week 9: July 31 - August 4
We spent this week on the poster and poster presentations. Our poster turned out very nice, and will go on a wall of the Parasol lab for the next group to better understand what we have done. Another part of this week was saying goodbye. It is amazing how many strong friendships can be built in such a short amount of time, but I hope we can all meet up again at conferences or maybe even grad school. The group here has been really great.
Since I came a week after the majority of participants, I will be at TAMU for one more week. In my last week I will be working on a paper with Harsh. Nancy wants this paper to be published when the project is complete.
Week 10: August 7 - August 11
The last week was quieter with the poster presentation done and a majority of the REU participants gone. Harsh and I spent this week perfecting our paper.
And the moral of the story is...
Research is fun! This has been a great learning opportunity. I had no idea about research in computer science before this summer. I am now 85% sure that I will go to Graduate School. Next summer I'd like to try out another institution through the DMP, but Texas A&M has a very welcoming program and I recomend them for any student looking for more experience in computer science.Last update 16 August 2006