Good design consists of finding the essential parameters and variables and avoiding unnecessary complexity. YAGNI (your aren’t gonna need it) principles As another example, consider a database for course grades: The essential information consists of names, students numbers, grades and comparable, but including details such as hair color, height and weight are irrelevant. Including color options or the interior have no role and should not be considered. Find what is essential to your problem, do not try to include everything possible (=avoid irrelevant details): If your task is to optimize the wind resistance of a new car model, its shape is the major factor. If your code fails to reproduce them, either your program, your algorithm or/and your general approach to solve it is/are wrong. In brief, the above is the essence of good design: Do your background work, solve the problem, find the important determinants and constraints, find analytical solutions that you can use as your benchmarks. Simplicity: The purpose is to solve problems, not to write the most beautiful code.įirst, solve the problem. Readability: A good code should include comments. Here, we will not focus on that issue, though. Parallelizability: A good code should be parallelizable. Portability: Codes should be written in such a way that they are portable, i.e., they must work correctly on different computer architectures In all programming, one should be careful and pay attention to detail, and remember debug, debug, and debug. It is impossible to write an exhaustive checklist, but the purpose of the short lists below is to bring some of the important issues to the reader’s attention. Independent of the chosen programming language, there are several matters that should be considered. Well-written programs: Desirable properties # The remarks below do not depend on the programming language. Plauger in The Elements of Programming Style.īefore we start discussing modelling and implementing the different algorithms and methods, let us briefly list a few general issues on design, programming, computing and codes.

Therefore, if you write the code as cleverly as possible, you are,īy definition, not smart enough to debug it. Review of thermodynamics and statistical mechanicsĭebugging is twice as hard as writing the code in the first place.

Module: Installation of necessary softwareĭigression: A few words about finite precision

Write a note on programming language