Xgboost: TPOT using XGBoost WARNING: /workspace/src/objective/regression_obj.cu:152: reg:linear is now deprecated in favor of reg:squarederror.
Hello,
I get this warning when using XGBoost in TPOT Pipeline, any idea why? TPOT developers say they do not set reg:linear.
fcoppey
All 4 comments
Fixed in master branch.
trivialfis
on 23 Jun 2019
When I install XGBoost on Mac with:
pip install xgboost
I'm keep getting this message. @trivialfis, can you suggest what to do to remove it?
vladimirkovacevic
on 3 Aug 2019
Use reg:squarederror or build from master branch. See document about objective parameter.
trivialfis
on 3 Aug 2019
I installed XGBoost Few days before and I was facing the same problem.
The answer is given in the error message. I changed Regressor model parameter: objective = 'reg:lineaqr' to 'reg:squarederror'.
example:
xg_reg = xgb.XGBRegressor(objective ='reg:squarederror', colsample_bytree = 0.3, learning_rate = 0.1, max_depth = 5, alpha = 10, n_estimators = 10)
And now its not giving any warning.
KhaaQ
on 7 Aug 2019
Related issues
FabHan
路
4Comments
yananchen1989
路
3Comments
hx364
路
3Comments
tqchen
路
4Comments
matthewmav
路
3Comments
Most helpful comment
I installed XGBoost Few days before and I was facing the same problem.
The answer is given in the error message. I changed Regressor model parameter: objective = 'reg:lineaqr' to 'reg:squarederror'.
example:
xg_reg = xgb.XGBRegressor(objective ='reg:squarederror', colsample_bytree = 0.3, learning_rate = 0.1, max_depth = 5, alpha = 10, n_estimators = 10)
And now its not giving any warning.