ARPACK and large eigenvalue problems

We have been working on large scale eigenvalue problems for sometime now. While i have been happy with the results from LAPACK, we wanted to achieve more insights on the problem by increasing the matrix dimension of our problem. However, increasing its dimension would bring us to great resources problem. Like most research group in the Philippines, we never get adequate funding from the government. Specially, since we are dealing with problems not really for commercialization. At least for now.

Dumping the proprietary MATLAB due to its costly license lead me to free and open source solutions. LAPACK is the first scientific package i used. Being new to GNU/Linux made it even harder to use such package. During those time, i have installed mandrake us an email server and a simulation workstation. Buhay pa naman yata ngayon iyon...

I have been using debian for more than three years now. The extent of our research is also growing. Our results are slowly being discussed by other research groups outside the country. We have also published our first Physical review letter article.

The above success lead me again to enhance more our previous results. For a week now, i am struggling to implement ARPACK to our problem. Today, i have successfully reproduce all previous results. ARPACK is a great tool. Its availablity from the apt repositories let me focus more on our solution rather configuring it. Also, i have been using SCILAB for a week now. I should have used long before. My regrets... hehe...

Anyways, i am feeling great relief today. thanks to ARPACK and the many debian developers for maintaining debian...

By the way, as my contribution the Debian GNU/Linux project, i joined its localization team. I am currenty running Debian with Tagalog messages... hehehe ... See below...