Logistic PLR Models

Coverage

The simulations are based on the the make_lplr_LZZ2020-DGP with \(500\) observations.

DoubleML Version                            0.11.5.dev55
Script                         LPLRATECoverageSimulation
Date                                    2026-08-10 12:55
Total Runtime (minutes)                       141.888912
Python Version                                    3.12.3
Config File              scripts/plm/lplr_ate_config.yml

Nuisance space

ⓘLearner m Learner M Learner t Bias CI Length Coverage
LGBM Regr. LGBM Clas. LGBM Regr. 0.146 0.693 0.944
LassoCV Logistic LassoCV 0.151 0.681 0.930
RF Regr. RF Clas. RF Regr. 0.097 0.452 0.934
Coverage for 95%-Confidence Interval over 500 Repetitions
ⓘLearner m Learner M Learner t Bias CI Length Coverage
LGBM Regr. LGBM Clas. LGBM Regr. 0.146 0.582 0.894
LassoCV Logistic LassoCV 0.151 0.571 0.870
RF Regr. RF Clas. RF Regr. 0.097 0.380 0.882
Coverage for 90%-Confidence Interval over 500 Repetitions

Instrument

ⓘLearner m Learner M Learner t Bias CI Length Coverage
LGBM Regr. LGBM Clas. LGBM Regr. 0.164 0.778 0.956
LassoCV Logistic LassoCV 0.165 0.703 0.916
RF Regr. RF Clas. RF Regr. 0.097 0.468 0.932
Coverage for 95%-Confidence Interval over 500 Repetitions
ⓘLearner m Learner M Learner t Bias CI Length Coverage
LGBM Regr. LGBM Clas. LGBM Regr. 0.164 0.653 0.910
LassoCV Logistic LassoCV 0.165 0.590 0.840
RF Regr. RF Clas. RF Regr. 0.097 0.393 0.900
Coverage for 90%-Confidence Interval over 500 Repetitions

Tuning

The simulations are based on the the make_lplr_LZZ2020-DGP with \(500\) observations. This is only an example as the untuned version just relies on the default configuration.

DoubleML Version                                    0.12.dev0
Script                        LPLRATETuningCoverageSimulation
Date                                         2025-12-01 14:39
Total Runtime (minutes)                             38.923752
Python Version                                         3.12.9
Config File              scripts/plm/lplr_ate_tune_config.yml

Nuisance space

ⓘLearner m Learner M Learner t Tuned Bias CI Length Coverage Loss M Loss a Loss m
LGBM Regr. LGBM Clas. LGBM Regr. False 0.198 0.966 0.940 0.727 0.355 0.479
LGBM Regr. LGBM Clas. LGBM Regr. True 0.202 0.812 0.910 0.631 0.331 0.469
Coverage for 95%-Confidence Interval over 100 Repetitions
ⓘLearner m Learner M Learner t Tuned Bias CI Length Coverage Loss M Loss a Loss m
LGBM Regr. LGBM Clas. LGBM Regr. False 0.198 0.811 0.890 0.727 0.355 0.479
LGBM Regr. LGBM Clas. LGBM Regr. True 0.202 0.682 0.870 0.631 0.331 0.469
Coverage for 90%-Confidence Interval over 100 Repetitions

Instrument

ⓘLearner m Learner M Learner t Tuned Bias CI Length Coverage Loss M Loss a Loss m
LGBM Regr. LGBM Clas. LGBM Regr. False 0.280 1.132 0.920 0.731 0.352 0.376
LGBM Regr. LGBM Clas. LGBM Regr. True 0.213 0.940 0.920 0.631 0.331 0.335
Coverage for 95%-Confidence Interval over 100 Repetitions
ⓘLearner m Learner M Learner t Tuned Bias CI Length Coverage Loss M Loss a Loss m
LGBM Regr. LGBM Clas. LGBM Regr. False 0.280 0.950 0.890 0.731 0.352 0.376
LGBM Regr. LGBM Clas. LGBM Regr. True 0.213 0.789 0.860 0.631 0.331 0.335
Coverage for 90%-Confidence Interval over 100 Repetitions