[Bug]: xgboost in R

[Forwardlee0]

Describe the bug

In R, when I set the learner to "regr.xgboost", I get the following error:
Error in y[] <- lapply(y, reallocate) :
ALTLIST classes must provide a Set_elt method [class: XGBAltrepPointerClass, pkg: xgboost]

My OS is macos and the xgboost package in R was installed via the R CMD INSTALL . instead of the CRAN.

Minimum reproducible code snippet

library(DoubleML)
library(mlr3)
library(mlr3learners)
library(data.table)
set.seed(2)
ml_g = lrn("regr.xgboost", nthread = 2)
ml_m = ml_g$clone()
obj_dml_data = make_plr_CCDDHNR2018(alpha = 0.5)
dml_plr_obj = DoubleMLPLR$new(obj_dml_data, ml_g, ml_m)
dml_plr_obj$fit()
dml_plr_obj$summary()

Expected Result

dml_plr_obj$fit()

Actual Result

Error in y[] <- lapply(y, reallocate) :
ALTLIST classes must provide a Set_elt method [class: XGBAltrepPointerClass, pkg: xgboost]

Versions

sessionInfo()
R version 4.4.2 (2024-10-31)
Platform: aarch64-apple-darwin20
Running under: macOS Sequoia 15.1.1

Matrix products: default
BLAS: /System/Library/Frameworks/Accelerate.framework/Versions/A/Frameworks/vecLib.framework/Versions/A/libBLAS.dylib
LAPACK: /Library/Frameworks/R.framework/Versions/4.4-arm64/Resources/lib/libRlapack.dylib; LAPACK version 3.12.0

locale:
[1] en_US.UTF-8/en_US.UTF-8/en_US.UTF-8/C/en_US.UTF-8/en_US.UTF-8

time zone: Asia/Shanghai
tzcode source: internal

attached base packages:
[1] stats graphics grDevices utils datasets methods base

loaded via a namespace (and not attached):
[1] compiler_4.4.2 tools_4.4.2

packageVersion('DoubleML')
[1] ‘1.0.1’
packageVersion('mlr3')
[1] ‘0.22.1’
packageVersion('xgboost')
[1] ‘2.1.2.1’

[PhilippBach]

Hi @Forwardlee0 ,

thanks! As you probably guessed, I cannot replicate it locally (Windows, using xgboost from CRAN) ..

Have you tried to update the mlr3 and mlr3learners packages recently? I am not sure, but they issued some new releases which also crashed some of our checks, so maybe it's worth a try?
What happens if you use the CRAN version of xgboost?

[Forwardlee0]

In Windows, there is no problem with the implementation of "regr.xgboost". The reason why I use xgboost from github instead of cran is that for xgboost of R in macos, xgboost from github can implement parallel computing, but cran's xgboost cannot. In addition, using xgboost from github alone for regression analysis is no problem. The above error occurs only when xgboost is implemented in doubleml.

[Forwardlee0]

All of the packages is the latest

[Forwardlee0]

I implement the following code, it works well:
library(mlr3)
library(mlr3learners)
library(data.table)

task = tsk("mtcars")
task$col_roles$target = "mpg"

learner = lrn("regr.xgboost")

learner$train(task)
prediction = learner$predict(task)
print(prediction)