*** REGRESSION ANALYSIS ***
Regression on sigmoid curve, in order to find the derivatives.

Question

*** REGRESSION ANALYSIS ***
Regression on sigmoid curve, in order to find the derivatives.

frostbite · Answer

The problem is kinda straight forward:
I want to find the so called "meltingpoint" $T_{M}$ which we define as the maximum of the derivatives. The situation is also shown in the attachment.
I got lots of data with the OD as a function of the temperature at a corresponding concentration.

How would you guys attack this problem?

frostbite · Answer

I was personally thinking in terms of fitting with a $n$'th polynomial and then differentiate it  $n-1$ times and set the derivative equal to zero and solve for the melting point.

frostbite · Answer

A very poor picture of my data is found in the following attachment.

joannablackwelder · Answer

Sorry, don't have the time to look at it right now. Gotta go. Maybe later.

frostbite · Answer

@abb0t I really need your help here.

abb0t · Answer

the derivative is a slope. (henceforth, the ratio of change in the value of the function to change in the independent variable.)

frostbite · Answer

Yea, but I wonder if it a somehow valid approximation, haven't done non-linear regression in a long time.

abb0t · Answer

In other words, best fit line....