linear_tree parameter not recognised by R version of LightGBM

[SpeckledJim2]

Description

Warning message: [LightGBM] [Warning] Unknown parameter: linear_tree

Reproducible example

library(lightgbm)
x <- matrix(data = sample(rnorm(100L), size = 100L), ncol = 1L)
y <- 2L * x + runif(nrow(x), 0L, 0.1)
dtrain <- lgb.Dataset(data = x, label = y)

lgb_parameters <- list(
objective = "regression"
, linear_tree = TRUE # this parameter isn't recognised by lightGBM
, verbose = -1L
, metric = "mse"
, seed = 0L
, num_leaves = 2L
, bagging_freq = 1L
, subsample = 1.0
)

bst_linear <- lgb.train(
data = dtrain
, nrounds = 10L
, params = lgb_parameters
, valids = list("train" = dtrain)
)

Environment info

sessionInfo()
R version 4.0.2 (2020-06-22)
Platform: x86_64-pc-linux-gnu (64-bit)
Running under: Ubuntu 18.04.5 LTS

Matrix products: default
BLAS: /usr/lib/x86_64-linux-gnu/openblas/libblas.so.3
LAPACK: /usr/lib/x86_64-linux-gnu/libopenblasp-r0.2.20.so

lightgbm_3.1.1 R6_2.5.0, installed from CRAN using install.packages('lightgbm')

Many thanks for creating lightgbm!

[btrotta]

@SpeckledJim2 Linear trees are not yet available in the CRAN release. As a workaround you can build the package from source, see: #3910 (comment)

[StrikerRUS]

@SpeckledJim2
You can subscribe to LightGBM repository releases

or to #3872 particularly

to be notified about updates.

[SpeckledJim2]

That's great, all installed and working, thank you!

I will raise another issue when I have created an example because when using predict() with predcontrib = TRUE the feature contributions no longer sum to prediction for the row if linear_tree = TRUE . I think the feature contributions aren't using the linear model in the leaves, they are still using the constant value, even when linear_tree = TRUE.

[StrikerRUS]

I will raise another issue

Could you please check this one #3998? I guess it is the same case.

[github-actions]

This issue has been automatically locked since there has not been any recent activity since it was closed. To start a new related discussion, open a new issue at https://github.com/microsoft/LightGBM/issues including a reference to this.