[R-package] CVBooster$reset_parameter() is broken

[jameslamb]

Description

lgb.cv() returns an object of class CVBooster. That class has a public method, $reset_parameter(), which is intended to be used to update the parameters of each Booster in CVBooster$boosters.

That method, which was added about 5 years ago (f3f2f5a), currently throws an error when called.

Error: object 'boosters' not found

Reproducible example

library(lightgbm)
data("agaricus.train")
dtrain <- lightgbm::lgb.Dataset(
    data = agaricus.train$data
    , label = agaricus.train$label
)
cv_bst <- lightgbm::lgb.cv(
    params = list(
        num_leaves = 8L
    )
    , nfold = 3L
    , nrounds = 10L
    , data = dtrain
    , obj = "binary"
    , verbose = -1L
)
cv_bst$reset_parameter(
    list(num_leaves = 11L)
)

Error in cv_bst$reset_parameter(list(num_leaves = 11L)) :
object 'boosters' not found

Environment info

LightGBM version or commit hash: v3.3.1.99 (7b10baf)

Command(s) you used to install LightGBM

sh build-cran-package.sh --no-build-vignettes && \
R CMD INSTALL ./lightgbm_3.3.1.99.tar.gz

Using R 4.1.0 on macOS.

Additional Comments

In the Python package, this works without issue.

import lightgbm as lgb
from sklearn.datasets import make_regression

X, y = make_regression(n_samples=10_000, n_features=10, n_informative=5)

dtrain = lgb.Dataset(data=X, label=y)

results = lgb.cv(
    train_set=dtrain,
    params={
        "objective": "regression_l2",
        "num_leaves": 8,
        "verbose": -1
    },
    stratified=False,
    shuffle=False,
    return_cvbooster=True
)
cv_bst = results["cvbooster"]
for bst in cv_bst.boosters:
    print(bst.params)

# {'objective': 'regression_l2', 'num_leaves': 11, 'verbose': -1, 'num_iterations': 100, 'early_stopping_round': None}
# {'objective': 'regression_l2', 'num_leaves': 11, 'verbose': -1, 'num_iterations': 100, 'early_stopping_round': None}
# {'objective': 'regression_l2', 'num_leaves': 11, 'verbose': -1, 'num_iterations': 100, 'early_stopping_round': None}
# {'objective': 'regression_l2', 'num_leaves': 11, 'verbose': -1, 'num_iterations': 100, 'early_stopping_round': None}
# {'objective': 'regression_l2', 'num_leaves': 11, 'verbose': -1, 'num_iterations': 100, 'early_stopping_round': None}

_ = cv_bst.reset_parameter({"num_leaves": 11})
for bst in cv_bst.boosters:
    print(bst.params)

# {'objective': 'regression_l2', 'num_leaves': 11, 'verbose': -1, 'num_iterations': 100, 'early_stopping_round': None}
# {'objective': 'regression_l2', 'num_leaves': 11, 'verbose': -1, 'num_iterations': 100, 'early_stopping_round': None}
# {'objective': 'regression_l2', 'num_leaves': 11, 'verbose': -1, 'num_iterations': 100, 'early_stopping_round': None}
# {'objective': 'regression_l2', 'num_leaves': 11, 'verbose': -1, 'num_iterations': 100, 'early_stopping_round': None}
# {'objective': 'regression_l2', 'num_leaves': 11, 'verbose': -1, 'num_iterations': 100, 'early_stopping_round': None}

[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.