sdm-help-list > How to do non-linear regression analyses?
Showing 1-3 of 3 posts
Display:
Results per page:
May 20, 2022  04:05 AM | Guochun Yang
How to do non-linear regression analyses?
Dear experts,

Thanks a lot for your great software.

I have conducted linear regressions with SDM and it really did a great job. Now I am considering to do some non-linear regression with SDM. For example,  I want to fit a gamma function curve to the data. Can I use SDM to achieve that? If not, is it possible to export the data to a generally readable format,  and run the regression with other software (such as Matlab or R)?

Any help will be appreciated.

Guochun
May 24, 2022  02:05 PM | Lydia Fortea - Instituto de Investigaciones Biomédicas August Pi i Sunyer (IDIBAPS)
RE: How to do non-linear regression analyses?
Dear Guochun, 

Unfortunately, SDM only conducts linear regressions. In order to be able to fit a non-linear regression you should do all the steps made in SDM to create the imputations, which is extremely hard and long. 

Sorry for not be able to help, 
Kind regards, 
Lydia
May 25, 2022  12:05 AM | Guochun Yang
RE: How to do non-linear regression analyses?
Hi Lydia,

Thanks a lot for your reply.

I have been able to run a non-linear regression by adding the quadratic of the linear predictor into the regression model in SDM, and it appears to have done a good job, since I plotted the results and the predicted values could largely fit the trend. Does this kind of analysis make sense? If so, I may consider adding a cubic term to fit more complex curves.

Best,
Guochun