I have created a partial dependence plot and converted a probability on the y axis. I am systematically going over my variables with the highest importance and see if there are some clear trends in predictor variables and the class. Since I am new to this, I would like to get some feedback from the scientific community.

I have a question mainly on bogs and tidal marsh. I noticed that the line disappears after a certain value on the x axis. Shouldn't the line go all the way to the end of the chart but be at around 0? Does it look like there are some errors in these charts?

enter image description here

  • $\begingroup$ Are there any points in your data for a Bog above 400 CHM mean (cm)? Many software packages stop plotting the line in a range where there is no data. Check the help page in your software. $\endgroup$
    – R Carnell
    Feb 12 at 15:29
  • $\begingroup$ Hi, nope there isn't. $\endgroup$ Feb 12 at 15:43
  • $\begingroup$ Figured out what I did wrong. I was having troubles making the PDPs from my rf model so I subdivded the dataset based on class then made the PDPs. Don't do that. I discovered the which.class argument and that solved my issues. $\endgroup$ Feb 15 at 2:48


Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Browse other questions tagged or ask your own question.