Polynomial approximation to find function values

Click For Summary

Discussion Overview

The discussion revolves around the use of polynomial approximation to determine the constants a and b in the equation R = a · e^(b/T) based on given temperature and resistance data. Participants explore methods for fitting the data and interpreting the results from MATLAB's polyfit function.

Discussion Character

  • Homework-related
  • Mathematical reasoning
  • Technical explanation

Main Points Raised

  • One participant describes the data and the equation R = a · e^(b/T), seeking to find constants a and b using polynomial approximation.
  • Another participant suggests plotting ln(R) against 1/T and using the output of MATLAB's polyfit to derive values for a and b.
  • A participant interprets the polynomial coefficients output by MATLAB and questions how to relate these to the constants a and b.
  • There is a suggestion to take logarithms of both sides of the equation to facilitate finding a and b from the slope and intercept of the plot.
  • One participant confirms the interpretation of the polynomial coefficients and seeks clarification on how they relate to the constants.

Areas of Agreement / Disagreement

Participants generally agree on the approach of using logarithmic transformation and polynomial fitting, but there are varying interpretations of the output from MATLAB and how to extract the constants a and b from it. The discussion remains unresolved regarding the exact method to determine b from the polynomial coefficients.

Contextual Notes

Participants express uncertainty about the relationship between the polynomial coefficients and the constants a and b, particularly in the context of series approximations. There are also assumptions about the accuracy of the coefficients and their significance in determining the constants.

Maxo
Messages
160
Reaction score
1

Homework Statement


If we have the following data

Code:
T = [296 301 306 309 320 333 341 349 353];
R = [143.1 116.3 98.5 88.9 62.5 43.7 35.1 29.2 27.2];
(where T = Temperature (K) and R = Reistance (Ω) and each temperature value corresponds to the resistor value at the same position)

Homework Equations


We know that
[tex]R=a \cdot e^{b/T}[/tex]
The question of the problem is the following: How can we, from these data and the given equation, find the values of the constants a and b?

The Attempt at a Solution


In the assignment it is suggested one draws a diagram that shows ln(R/Ω) as a function of 1/T and then use MATLAB "polyfit" function which gives a polynom approximation of this line and then use that polynom to find the values of a and b.

I have used the polyfit function in MATLAB to get a polynom approximation, but how does that polynom give me the constants a and b?
 
Last edited:
Physics news on Phys.org
Polyfit outputs polynomial coefficients, so I think what you'd have to do is plot ln(R) = ln(a) + b/T, and look for b = 0. Now you will know what value a is. From there, divide your polyfit matrix by a, and you will have the coefficients to the power series of eb/T. With the power series representation, you may have to do some approximating to find a value for b, but I would imagine the coefficients won't change much after the second order term.
 
So MATLAB gives me this polynomial:

Code:
1.0e+03 * 

3.0412   -0.0053
which I assume means the same as a polynomial like
[tex]y=10^{3}\cdot 3.0412x-0.0053 = 3041.2x - 5.3[/tex]
I assume then, that -5.3 is the value of ln(a)?

If I divide then the polynomial with -5.3 I get the following matrix:
Code:
[-0.0018    1.0000]
Correct so far?

If so I wonder what does it means that these are the coefficients for the power series of e^(b/T)? I guess that should be a Maclaurin or Taylor series? How can I determine b from such a series?
 
We know that
R=a⋅eb/T

The question of the problem is the following: How can we, from these data and the given equation, find the values of the constants a and b?
If I were in your position, I would take logarithms of both sides of that equation, and plot a graph of your data. The y-intercept and the slope lead you to values for a and b. You will then know the approximate values for those constants, and can recognize when MATLAB is giving you nonsense.
 
Maxo said:
I assume then, that -5.3 is the value of ln(a)?
Right (I'd go with at least 3 significant figures).

You'd like to fit your data to:
[tex] \ln(R) = b \frac{1}{T} + \ln(a)[/tex]
and MATLAB has provided you with:
[tex] \ln(R_\mathrm{fit}) = p_1 \frac{1}{T} + p_2[/tex]
You're basically done at this point. How does ##p_1## and ##p_2## relate to ##b## and ##a##, respectively?
 
Thanks, that's how I figured also. Onwards
 

Similar threads

Replies
2
Views
2K
Replies
4
Views
10K
  • · Replies 7 ·
Replies
7
Views
5K
  • · Replies 5 ·
Replies
5
Views
2K
  • · Replies 3 ·
Replies
3
Views
3K
Replies
1
Views
1K
  • · Replies 46 ·
2
Replies
46
Views
9K
Replies
18
Views
5K
Replies
1
Views
2K
  • · Replies 37 ·
2
Replies
37
Views
6K