Numerical Error Propagation
Contents |
on a series designed to survey the design and analysis of various numerical methods will look at error propagation. 28 May, 2013 - Article, Numerical Analysis - Tags : error propagation, floating point, numerical analysis 3 Comments Tweet
Types Of Error In Numerical Analysis
Table Of Content 1. Safety First 2. A Modern Day Little Gauss Story At propagation error a Brief Glance 3. Some Basics - Errors Error Propagation Arbitrary Differentiable Function Table of Error Propagation 4. Example 1. 2.
Error Propagation Example
3. There was recently a good article on scientific computing, defined loosely as the dark art, as it may have seemed to the uninitiated, of deriving solutions to equations, dynamical systems, or what-not that would have made your Mechanics professor scream in horror at the thought that they will need to somehow "solve" these systems. Of course, in the rich computer world today, almost any problem imaginable (exaggeration of course!) can already be solved by some existing tool. Hence more and more, the focus gets shifted from "how do I solve this differential equation" to "what do I ask google?" My dad once told me of a glorious time "back in [his] days" when every respectable engineering institution would have a crash course on this dark art on top of Fortran. Of course, my dad is only 43, and that was only 19 years ago. Even now, when computer science departments everywhere no longer believes in the necessity in forcing all of their graduates to have a basic grasp on numerical analysis, there is still some draw in the subject that either makes people deathly afraid of it or embrace it as their life ambition. I am personally deathly afraid of it. Even then, there are quite many cute gems in the field, and as such, I am still very much so attracted to the field. Scientific computing is the all encompassing field involving the design and analysis of numerical methods. I intend to start a survey of some of the basic (but also most useful) tools such as methods that: solve linear and nonlinear systems of equations, interpolate data, compute integrals, and solve differential equations. We will often do this on problems for which there exists no "analytical" solution (in terms of the common transcendental functions that we're all used to). 1. Safety First In an ideal world, there would be a direct correspondence between numerical algorithms their implementation.
numbers to powers, etc. can be determined with specific analytical formulas. However, the analysis of the propagation of errors through a model is frequently most easily http://www.phailed.me/2013/05/introduction-to-scientific-computing-error-propagation/ accomplished numerically. Analytical Methods Formulas for functions of one variable Formulas for functions of two variables « Previous Page Next Page » Quantitative Skills Issues and Discussion Teaching http://serc.carleton.edu/quantskills/teaching_methods/und_uncertainty/errpropagation.html Methods Back of the Envelope Calculations Mathematical and Statistical Models Measurement and Uncertainty Metacognition Models Teaching Quantitative Literacy Teaching Quantitative Reasoning with the News Teaching with Data Teaching with Data Simulations Teaching with Equations Teaching with SSAC Understanding Uncertainty Appropriate Representation of Numbers Significant Figures Rounding Numbers Propagation of Error Precision and Accuracy Measurement Error Teaching Resources Settings Tools and Datasets Community Last Modified: June 21, 2012 | Printing | Shortcut: http://serc.carleton.edu/37886 | Privacy | Terms of Use | Report a Problem/Feedback
be down. Please try the request again. Your cache administrator is webmaster. Generated Sat, 22 Oct 2016 04:44:46 GMT by s_wx1126 (squid/3.5.20)