topic Re: Add linear fit , eqn, R^2 to overlay plots in Discussions

Add linear fit , eqn, R^2 to overlay plots

ad — Thu, 21 Jun 2018 17:43:47 GMT

Hi,

I cannot add linear fit and display equation & R^2 values when using overlay plot feature in jmp by label : 0 and 1. See the attached excel for data and desired plot.

Can you please help ?

Thanks a lot.

Re: Add linear fit , eqn, R^2 to overlay plots

Dan_Obermiller — Thu, 21 Jun 2018 17:49:48 GMT

Your Excel sheet does not have the data, but here is an approach that should work.

Go to Graph Builder.

Drag your X variable to the X zone.

Drag your Y variable to the Y zone.

Drag your Label variable to the Overlay zone.

This should give you a scatterplot with different colors for each value of the Label variable.

Now click on the Graph element for the regression line.

You can then modify what is displayed by looking at the regression line options to the left of the graph.

Here is an example using the Big Class data.

Re: Add linear fit , eqn, R^2 to overlay plots

ad — Thu, 21 Jun 2018 18:43:37 GMT

Hi Dan,

Thanks a lot for your quick solution. I appreciate it.

One more thing: If I want to change Y axis to log scale & want to do Fit Special : Transformed Fit log, can I acheive this in Graph Builder mode ?

THanks,

Re: Add linear fit , eqn, R^2 to overlay plots

Dan_Obermiller — Fri, 22 Jun 2018 17:10:44 GMT

Graph Builder is designed for graphs, not analysis. Fortunately, everything you want to do can be done through Fit Y by X which is an analysis platform.

Choose Fit Y by X. Specify your X and Y variables.

When the scatterplot appears, go to the red triangle and choose Group By... and specify your labels variable.

Go to Fit Special and choose your transformed fit. It will be done for each value of the labels variable.

Right-click the Y-axis and select Axis Settings. In the dialog, in the upper left is the scale control for you to change to a log axis.

Re: Add linear fit , eqn, R^2 to overlay plots

ad — Fri, 22 Jun 2018 16:56:11 GMT

Hi Dan,

Thanks a lot! Really appreciate your help.