Help understanding parameters of an exponential model

[kaska5]

Hello

The aim is to fit a model:

y = a * exp( - b * X) + c

to log-transformed data. X is the independent variable, not log-transformed.
Visually, the data obey a power-law with a negative exponential decrease.

How can I get the slope (m) and intercept (b) as I would in a normal regression model:

y = m*X + b

Any help is appreciated.

 

[Steven G]

(y-c)/a = exp(-bX). Can you solve for X now?

What is this y=mX+b all about? Do you want to write the original equation in this form?

In the future you should obey the sites guidelines which states that to receive help you need to show your work so the helpers know what you need help with.

 

[JeffM]

(y-c)/a = exp(-bX). Can you solve for X now?

What is this y=mX+b all about? Do you want to write the original equation in this form?

In the future you should obey the sites guidelines which states that to receive help you need to show your work so the helpers know what you need help with.

Jomo

The question is not particularly cohereht, but I suspect what is wanted is to use linear refression software to fit a line of best fit to what is presumed to be an exponential relationship.

 

[Steven G]

Jomo

The question is not particularly cohereht, but I suspect what is wanted is to use linear refression software to fit a line of best fit to what is presumed to be an exponential relationship.

Yeah, I misread that one!

 

[HallsofIvy]

"refression software"? There is software to refresh you drink now?

 

[JeffM]

"refression software"? There is software to refresh you drink now?

Yes. It is part of the internet of things.