physics1_lab

code (and perhaps results) useful for the physics1_lab in the Technion

error analysis

The file error_calc.m is a matlab class that assits with error analysis and error propagation. The usage explains it much better than I can. All errors are assumed to be non systematic (i.e. random)

A small disclaimer

error_calc should not be used where systematic errors are involved, only where random errors occur. Support for systematic errors is not yet implemented, though it is planned

usage

An example is the best way to understnd the usage (and also motiviation)

l = 127;
l_err = 1;
h = 2.5;
h_err = 0.1;
% construct class objects
l = error_calc(l,l_err) / error_calc.accu(1000);
h = error_calc(h,h_err) / error_calc.accu(1000);

% in prompt
>> h/l

ans = 

  error_calc with properties:

    meas: 0.0197
     err: 8.0251e-04

>> h+l

ans = 

  error_calc with properties:

    meas: 0.1295
     err: 0.0010

Note the difference between creating values with error (just constructing error_calc) and creating values which are assumed to be accurate (for example, meter to milimeter conversion, multiplication by a number etc..).

Theoretical basis

All equations are taken/derived from:

• Taylor, J.R. (1997) An Introduction to Error Analysis: The Study of Uncertainties in Physical Measurements. 2nd Edition, University Science Books, Sausalito.

Specifically, chapter 3, where formulas for error propagation of random errors are done.

tasks and improvements for the future

• Add more capabilities to error_calc (such as sin function, getting a derivative in a point etc..)
• De-dplucating code for the various ways matlab does division and multiplication
• Add support for systematic errors, i.e. implement different structs/ different operators such that using them uses a different formula for errors