topic LOWESS (Locally weighted regression and smoothing scatterplots) in Discussions

LOWESS (Locally weighted regression and smoothing scatterplots)

Teresa_ — Tue, 19 Nov 2019 13:48:38 GMT

Hey all,

I am new to JMP and try currently my very best :)

My problem is that I want to do a LOWESS with my data and want to further work with the received equation.

Therefore I can not use the Smoother in Graph Builder.

I already got help from the support. I should put a formula for LOWESS into a column and then use this for analysis under Specialized Modeling / Nonlinear. I already have found the equation (https://www.weisang.com/en/documentation/loessandlowessalgorithm_en/#). However, I have a problem with entering the formula.

Is there somebody who could give me some advise on how to best set this up?

Thank you in advance for your help.

Re: LOWESS (Locally weighted regression and smoothing scatterplots)

David_Burnham — Tue, 19 Nov 2019 14:43:54 GMT

From the ANALYZE menu choose Fit Y by X and specify your x and y coluimns. This will take you into the BIVARIATE platform.

From the red traingle select FLEXIBLE> KERNEL SMOOTHER.

That will apply a smoothing curve based on what I believe is the LOWESS method. You can use the slider to adjust the stiffness of the curve.

From the red triangle next to "Local Smoother" select SAVE PREDICTION FORMULA.

Hey presto, in your data table you now have the formula for the LOWESS (LOESS) curve.

Re: LOWESS (Locally weighted regression and smoothing scatterplots)

Teresa_ — Wed, 20 Nov 2019 07:14:13 GMT

Hey David, thank you for your quick response. This was really helpful! :)