Dismiss Notice
Join Physics Forums Today!
The friendliest, high quality science and math community on the planet! Everyone who loves science is here!

Calculating uncertainty

  1. Apr 15, 2008 #1
    I have a lab where I have to calculate the theoretical value of R using the following equation and then find the uncertainty in R.

    *I know the values of x,y,z and their respective uncertainties.

    The problem is that we have only learned basic uncertainty rules (i.e. for multiplication/division you add the %uncertainty, for addition/subtraction you add the absolute uncertainties). This is much more complicated since I have to deal with SIN/COS and square roots. I was searching around and it seems that I have to calculate the partial derivative or differentials. I am not familiar with differentials and I have no idea how to solve this problem. If anyone can offer any help whatsoever it would be greatly appreciated or anywhere where I can find this information.
  2. jcsd
  3. Apr 15, 2008 #2
    I'm using excel if that makes any difference whatsoever
  4. Apr 17, 2008 #3
    You are quite right about using partial derivatives. Don't be scared yet. I agree that you have a complicated expression to start with. I would suggest first simplifying it as much as possible (using trig identities). I assume you know basic differentiation? I really hope so...
    When I say "simplify" I mean put it in a form thats less "scary" to differentiate. (derivatives of sin and cos can be found in any table of derivatives).

    Example: If you have a function [tex] T=T(f,\lambda,...) [/tex] then the error in T is found by:
    [tex] (\delta T)^2 = (\frac{dT}{df})^2(\delta f)^2 + (\frac{dT}{d\lambda})^2(\delta \lambda)^2+(\frac{dT}{d...})^2(\delta ...)^2 [/tex]
    where [tex] \delta [/tex] is the error value(s).
  5. Apr 17, 2008 #4
    I don't know if Excel can differentiate...Ever used Matlab or Mathematica?
  6. Apr 17, 2008 #5
    To calculate the partial derivative [itex]\partial R/\partial x[/itex], differentiate R with respect to x, treating all the other variables as constants. Do likewise for the other partial derivatives.
    Last edited: Apr 17, 2008
Share this great discussion with others via Reddit, Google+, Twitter, or Facebook