Skip to content

question about predict using lme4

2 messages · Li, Xiaoma [GE AT], Lionel

#
Hi all,
I want to use lme4 to predict air temperature using land surface temperature.
I now can ran the package successfully. But I am not familiar to the parameters of the predict function. Here is my code: plsttest=predict(mod.xm,lsttest2,allow.new.levels = T)
I have two questions: Is there some parameters missed?
How does lme4 predict new levels (new climate station in my study?) Does it use fixed model or others?

Thank you very much!

Sincerely
Xiaoma li
#
Dear Xiaoma Li,

To answer your questions:

- No parameters are missed (why should they?)

- If you have in your lsttest2 data frame new levels for the random 
factors, new random deviates will be generated for these levels from a 
normal distribution with mean 0 and standard deviation equal to the 
estimated random standard deviation (to get these estimated values look 
at the summary output).

Hope this helps.

Yours,
Lionel
On 05/05/2016 15:46, Li, Xiaoma [GE AT] wrote: