topic Re: How to lower the plateau of a 4 parameters fit curve? in Discussions

How to lower the plateau of a 4 parameters fit curve?

aschmitz1 — Thu, 02 May 2024 15:39:05 GMT

Hello !

I'm working on dose-response curves. I use 4-parameter logistic curves from specialised modeling. My first plateau doesn't fit properly, it's too high, because my first point is too high. I don't know what parameter to set and how to set it so that my plateau goes down and gets closer to 0.

Thank you in advance for your answer!

Re: How to lower the plateau of a 4 parameters fit curve?

Ross_Metusalem — Fri, 03 May 2024 20:51:26 GMT

You might consider using the Nonlinear platform instead of Fit Curve. Both fit nonlinear models, but Nonlinear provides more detailed options for controlling parameter estimation, among other things. (Fit Curve is for simplified out-of-the box nonlinear modeling.) One of the Red Triangle options in Nonlinear is declaring lower or upper bounds on parameter estimates, which you can use to address an asymptote being too high or low.

Note that Nonlinear requires a model equation specified as a column formula in your data table. It can write one for you if you're using one of JMP's built-in models. In the launch dialog, just use the Model Library button to declare a Logistic 4P model to get started.

Re: How to lower the plateau of a 4 parameters fit curve?

Byron_JMP — Mon, 06 May 2024 18:40:58 GMT

another option is to scale your response variable 0-1.