DataMonkey is a C library containing implementations of various algorithms that I found to be utilitarian, performant, and just down right interesting. The different structures are being coded as I progress through my understanding of C, Linux, Distributed algorithms, and my general courses of study at Califorinia State University, Sacramento (csus).
- So that I can understand basic concepts of algorithm design. Reading in a book or paper, or even worse just blindly using an abstract api only goes so far in the knowledge that I can attain. Plus I learn better with a more "hands on" approach.
- Create cutting edge algorithms. It would be terribly difficult for me to hack up a fibonacci heap, without a solid understanding of a basic heap or tree structure first. The idea is to gradually build up my understanding to newer and more cutting edge algorithms. The approach taken following something along the lines of. . .
- Learn algorithm or novel technique or data structure, etc.
- Layout and create the initial idea.
- Tests basic routines.
- Profile, to see if more complexity warranted.
- Learn alternate approaches if necessary.
- And repeat with something more advanced.
- Quality comments along with tips for my forgetful self
- "If I have to think longer than 30 seconds, comment it."
- Test for edges cases from multiple directions.
- Standardize look of code by a syntax formatter "astyle".
- Keep an up to date wiki on progress, api's, etc.
- Understand that it's outside or your daily course work and job, just don't overcommit yourself.
- Keep to these project guidelines. :)