Compoutational Physics
- 1. Computational Physics By Saadia Shaukat
- 2. Computational Physics Computational physics is the study and implementation of numerical analysis to solve problems in physics for which a quantitative theory already exists. It is sometimes regarded as a subdiscipline of theoretical physics. On the other hand, consider it is considered an intermediate branch between theoretical and experimental physics - an area of study which supplements both theory and experiment.
- 3. Computational Physics A representation of the multidisciplinary nature of computational physics both as an overlap of physics, applied mathematics, and computer science and as a bridge among them
- 4. Overview In physics, different theories based on mathematical models provide very precise predictions on how systems behave. Limits of mathematical models When the solution does not have a closed-form expression. When the solution is too complicated. In such cases, numerical approximations are required.
- 5. How Computational Physics works The solution is approximated by the following steps: 1. Write the problem as a finite number of simple mathematical operations (algorithm). 2. Use Computer to perform these operations and compute an approximated solution and respective error.
- 6. Challenges Computational physics problems are in general very difficult to solve exactly. This is due to several (mathematical) reasons: 1. lack of algebraic and/or analytic solubility 2. Complexity 3. Chaos In addition, the computational cost and computational complexity for many- body problems tend to grow quickly. Finally, many physical systems are inherently nonlinear at best, and at worst chaotic: this means it can be difficult to ensure any numerical errors do not grow to the point of rendering the 'solution' useless.
- 7. Computational Science The focus of computational science is how to use the computer to solve scientific problems. computational scientist aims to develop stable, portable and self-consistent algorithms, all within the context of the problem being solved. Computational physics is the application of computational science to physics problems.
- 8. Computational Physics More specifically Computational physics is the study and implementation of numerical algorithms to solve problems in physics by means of computers. Computational physics in particular solves equations numerically. Finding a solution numerically is useful, as there are very few systems for which an analytical solution is known. Another field of computational physics is the simulation of many- body/particle systems; in this area, a virtual reality is created which is sometimes also referred to as the 3rd branch of physics (between experiments and theory).
- 9. Computational Physics (cont) The evaluation and visualization of large data sets, which can come from numerical simulations or experimental data (for example maps in geophysics) is also part of computational physics. Another area in which computers are used in physics is the control of experiments.
- 10. Importance of computational Physics Computational physics plays an important role in the following fields: Computational Fluid Dynamics (CFD): solve and analyze problems that Involved in fluid flows Solid State Physics (Quantum Mechanics) High Energy Physics / Particle Physics: in particular Lattice Quantum Chromodynamics (Lattice QCD) Astrophysics: many-body simulations of stars, galaxies etc. Geophysics and Solid Mechanics: earthquake simulations, fracture, rupture, crack propagation etc. Agent Models (interdisciplinary): complex networks in biology, economy,social sciences and many others
- 11. Computing Platforms Computational physicists need to be able to work on a variety of different computing platforms. Platforms can be categorized in two ways: by the type of CPU installed, and by the operating system used. These two are interconnected, but since there are a large number of different CPUs, it is easier to look at operating systems. There are three main operating systems used in the world today: Unix based, Mac-OS based, and MS Windows based. Each has its own strengths and weaknesses.
- 12. Programming Languages Just as there are a number of different computing platforms, there are also a wide variety of programming languages available. The languages are typically classified according to the advances that they incorporate, with each major advance represented as a new generation.
- 13. First Generation Language The first generation language is machine code, with the program written as a string of numbers that the computer interpreted as instructions. This was the most difficult way to program a computer, but is the closest to how the computer actually works. In fact, all later generation languages are translated into machine code before the CPU processes them.
- 14. Second generation languages Second generation languages are considered to be various assembler languages. Since they use English acronyms, these languages are easier to understand than straight machine code, but they are still arcane compared to later generation languages.
- 15. Third generation languages- High Level Languages Third generation languages are also known as “high level” languages. A compiler takes the statements of a particular third generation language and converts it into machine code.
- 16. Fourth generation languages Fourth generation languages are similar to third generation languages in that they rely on a compiler or interpreter, but they are usually written in a language which has structure and flow similar to spoken languages. In general, most computational physicists need expertise in third, or later, generation languages.
- 17. The main exception to this is if they are working on software that must run in an embedded system or in a real time operating system, and the speed of the code is paramount(supreme). In this case, the ability to trim any excess commands out of the code stream becomes important, and assembly languages are the best choice.
- 18. Of the high level languages, the one that are most commonly found in physics are Fortran, C/C++, MatLab, Mathematica, Maple, and PAW(Physics analysis work station).
- 19. Fortran Fortran is the oldest of the scientific computing languages. Originally designed in 1954, it went through various modifications. Fortran II was released in 1958. Fortran III was a short-lived release, Fortran IV, which was released in 1962, became the mainstay of scientific computing. A large number of scientific programs and libraries were written in Fortran IV. (Fortran 77) was released in 1977, these programs and libraries were quickly updated and augmented.
- 20. What made Fortran so widely accepted in the scientific community? The language had a number of built in data types that eased scientific computing, including real, double and complex. It also had a wide variety of controlling and branching structures. These two facts enabled large programs to be created which had the ability to track variables associated with experimental measurements.
- 21. Fortran 77 remained the standard scientific language for more than ten years. During this time computer science made a number of significant advances, prime among them the invention of pointers that allow indirect access to data. Various vendors released patches to Fortran 77 that allowed it to also do this, but it wasn’t until the release of Fortran 90 that Fortran gained many of these tools.
- 22. C/C++ C is a more recent entry into the field of scientific programming. It was released in 1971 and was designed to provide a computing development tool. Its strengths include the facts that It has high level constructs, It can handle low-level activities, and It produces efficient programs
- 23. From a scientific standpoint, the loose typing that C provided quickly enabled all of the strengths of Fortran to be copied and expanded upon. The rise of Unix workstations, whose kernel is written in C, also aided in the acceptance of C in the scientific community. In addition, the use of header files and longer variable names made documentation of code libraries easier.
- 24. As computer science paradigms advanced, the idea of object- oriented programming was developed. Object-oriented programming (OOP) uses a different approach to problem solving than the method used in older languages such as Fortran and C. These earlier languages use a method known as a “procedural approach”. A problem is broken into smaller problems, and this process is repeated until the subtasks can be coded. Thus a library of functions is created, communicating through arguments and variables
- 25. PAW, Mathematica and Maple IDL, PAW, Mathematica, and Maple are all examples of fourth generation languages. Unlike the third generation languages discussed above, these are interpreted languages. Interpreted languages are ones in which the code is passed directly to a parser, which converts it to machine code and executes it immediately. Interpreted languages do not usually produce a stand- alone executable, and are typically slower than their compiled counterparts.
- 26. ➢ IDL, which stands for Interactive Data Language, and PAW, which stands for Physics Analysis Workbench, are data analysis and graphics packages. ➢ They can read ASCII or binary data files, do a wide range of sophisticated mathematical operations on the data, and then present the results in either a textual or graphical output. ➢ Both packages are widely used as post processing software in experimental support.
- 27. In contrast, Mathematica and Maple are symbolic computation programs. ➢ While they can both carry out the same duties as IDL and PAW, their greatest strength lies in their ability to perform mathematical calculations using formulae similar to those that would be used in an analytical approach. ➢ For example, a user could input a set of linked equations and ask the program to solve them for a specific variable. The result will be presented (if possible) as an analytical formula rather than as a number. ➢ The fact that all of these packages are fourth generation languages makestheir use much more intuitive than standard “programming” languages. ➢ Users do not necessarily need to know arcane calls or coding techniques to achieve their goals.
- 28. Solving Equations One-Dimensional Case The problem of finding a root of an equation can be written as F(x)=0 Bisection Method
- 31. Matlab Implementation function p=bisec(a,b,TOL,N0,f) %syms x i=1; %f=x^2-6*x+8; FA=subs(f,a); found=0; while i<=N0 p=a+(b-a)/2; FP=subs(f,p); if FP==0 | (b-a)/2 <TOL
- 32. break; end i=i+1; if FP*FA >0 a=p; FA=FP; else b=p; end end
- 35. Example