×
INTELLIGENT WORK FORUMS
FOR ENGINEERING PROFESSIONALS

Log In

Come Join Us!

Are you an
Engineering professional?
Join Eng-Tips Forums!
  • Talk With Other Members
  • Be Notified Of Responses
    To Your Posts
  • Keyword Search
  • One-Click Access To Your
    Favorite Forums
  • Automated Signatures
    On Your Posts
  • Best Of All, It's Free!
  • Students Click Here

*Eng-Tips's functionality depends on members receiving e-mail. By joining you are opting in to receive e-mail.

Posting Guidelines

Promoting, selling, recruiting, coursework and thesis posting is forbidden.

Students Click Here

Jobs

Exponential rise fit

Exponential rise fit

Exponential rise fit

(OP)
An exponential decay ( y = C*e^(B*x) ) is straightforward to fit by taking the LN(y) and using LINEST in Excel.  But does anybody have a good way to fit the equation y = A - C*e^(B*x), where x and y are data points and A, B, and C are all unknown?

 

RE: Exponential rise fit

For any function you want to fit to data use the Solver tool to minimise the sum of the square of the differences between the actual and predicted Y value.

Have a column of data with initial estimates of A,B, and C and, using the function you want to fit, have a column of the square of the difference between the estimated function and the Y data values, with the total of these squared differences in another cell. Use the Solver to minimise that cell of the sum using the cells of A,B,C as the parameters to vary. You might have to play around with the initial estimates of A,B and C to get the best fit.  

Tara

RE: Exponential rise fit

Yi = A - C*e^(B*Xi)

You have an estimate of A, it is simply the average of the measurements Yi.

You can build a new random variable Y1i as follows:
Y1i = -1 * (Yi – Average{Yi} )

Then you can solve C and B using your old method based on:
Y1i = C*e*(B*Xi).

BUT, that does impose a constraint that the actual mean is exactly equal to your estimated mean, when in fact there is variability that must be apportioned to all the unknowns model parameters, including the mean.  So, you will get a lower residual sum of squares (better fit) if you use the method suggested by Corus.
 

=====================================
(2B)+(2B)'  ?

RE: Exponential rise fit

Sorry, I was way off base. The method I proposed assumed the mean of the 2nd term was zero, when there was no basis for that.

Never mind...

=====================================
(2B)+(2B)'  ?

RE: Exponential rise fit

(OP)
Thanks corus, I think your solution should work for me.

RE: Exponential rise fit

why not convert your "y" to "y-A" ?

instead of ln(y) linear fit to ln(y-A), choose A to maximise the r^2 term ??

Red Flag This Post

Please let us know here why this post is inappropriate. Reasons such as off-topic, duplicates, flames, illegal, vulgar, or students posting their homework.

Red Flag Submitted

Thank you for helping keep Eng-Tips Forums free from inappropriate posts.
The Eng-Tips staff will check this out and take appropriate action.

Reply To This Thread

Posting in the Eng-Tips forums is a member-only feature.

Click Here to join Eng-Tips and talk with other members!


Resources