landmap package, predict method gave an error of memory exhausted

Dear list members,

I am fitting an Ensemble Machine Learning with the R package landmap, using
the function train.spLearner, the resulting object is large 2.5 GB, and
when I tried to predict I I got the message of memory exhausted.

I have only 14 points and 23 covariables, in a very small area.

Using R 3.6.1
MacOS, 16GB of ram


pred_co <- predict(m_co)

Error: vector memory exhausted (limit reached?)
Error during wrapup: vector memory exhausted (limit reached?)

Manuel,

Please send me a private message and a copy of your data so I can test
where does the RAM blows up. Is it a classification or regression type
problem?

T. Hengl

On 9/16/19 4:01 PM, Manuel Sp?nola wrote:

Dear list members,

I am fitting an Ensemble Machine Learning with the R package landmap,

using

the function train.spLearner, the resulting object is large 2.5 GB, and
when I tried to predict I I got the message of memory exhausted.

I have only 14 points and 23 covariables, in a very small area.

Using R 3.6.1
MacOS, 16GB of ram


pred_co <- predict(m_co)

Error: vector memory exhausted (limit reached?)
Error during wrapup: vector memory exhausted (limit reached?)