View Single Post
  #1   Report Post  
teh2heartedale teh2heartedale is offline
Junior Member
 
Posts: 1
Default Fit a Logistic Curve

Hello all

First time poster, so I apologize for any formatting issues.

I am working on an assignment, and am posting here because I couldnt make it to class to learn the material.

I was given very simple data with X ranging from 1-5, and Y ranging from .4-.9.
x y
1 0.4016458
2 0.5747817
3 0.7313313
4 0.8457159
5 0.9169331

How do I fit a logistic curve f(x) = 1/(1+e^(-a*(x-c))) to this made up data?

plotting these points gives me a somewhat linear relationship but the instructions say that it is 1) not the same as a log curve, and 2) that i cant simply do this using a trendline.

What does he mean when he says that i require setting up decision variables, predicted values, residuals, and a objective function cell.

Iam suppose to use the Solver add-in.


== My questions ==
How do I identify the variable a and c?
Do I attain a, c and my predicted values from using the Analysis function over the y variable?

I am really confused about this topic. Ive been searching for "fit logistic curve excel" on google and youtube and many of these descriptions do not come close to what I am asking (i think). Usually i find some guidance in these help tutorials, however this time I can not find anything.


EDIT: I have been following a template that my prof created for a separate data and I have used Sine and Cosine variables to conduct a regression analysis to find predicted Y values. I got the prediction/residual of that data set, but am stuck as to fit the graph :(. Iam guessing that the sine and cosine were the decision variable, and if iam correct i do not know how to translate the concept to this problem.

Any guidance would be wonderful

Much Appreciated
teh2heartedale

Last edited by teh2heartedale : May 21st 13 at 05:49 AM Reason: Anecdotal supplement.