Skip to content
New issue

Have a question about this project? # for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “#”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? # to your account

Confusion about the prediction of β parameter in universal kriging #120

Open
Bamboobill opened this issue Feb 20, 2023 · 1 comment
Open

Comments

@Bamboobill
Copy link

I have a question about how to obtain a map of regression coefficients that varies with space when I do Multiple Linear Regression in universal kriging. For example, a linear regression of Y = k1X1+k2X2. Could I predict a spatial pattern of parameter k1 in gstat package?

I noticed that Liang et al (2016) used gstat package to estimate a global map of effects of species diversity on productivity, which means that the regression parameter (e.g., k1 and k2) varies with spatial location. However, I do not know how to implement this, though I have tried many times. Is there any function to get regression coefficients that varies with space in gstat package? Many thanks!

Reference:
Jingjing Liang. et al., 2016, Science 'Positive biodiversity-productivity relationship predominant in global forests'

@edzer
Copy link
Member

edzer commented Feb 20, 2023

Yes, but you'll have to fake it by providing a newdata object that has X1 values equal to 1 and X2 values equal to 0.

# for free to join this conversation on GitHub. Already have an account? # to comment
Labels
None yet
Projects
None yet
Development

No branches or pull requests

2 participants