IIVM Models

Coverage

The simulations are based on the the make_iivm_data-DGP with \(500\) observations. Due to the linearity of the DGP, Lasso and Logit Regression are nearly optimal choices for the nuisance estimation.

DoubleML Version                                0.12.dev0
Script                         IIVMLATECoverageSimulation
Date                                     2025-12-04 17:32
Total Runtime (minutes)                         22.869692
Python Version                                     3.12.3
Config File              scripts/irm/iivm_late_config.yml
ⓘLearner g Learner m Learner r Bias CI Length Coverage
LGBM Regr. LGBM Clas. LGBM Clas. 0.247 1.325 0.966
LGBM Regr. LGBM Clas. Logistic 0.245 1.319 0.974
LGBM Regr. Logistic LGBM Clas. 0.243 1.259 0.968
LGBM Regr. Logistic Logistic 0.239 1.252 0.975
LassoCV LGBM Clas. LGBM Clas. 0.232 1.253 0.974
LassoCV LGBM Clas. Logistic 0.229 1.247 0.982
LassoCV Logistic LGBM Clas. 0.224 1.188 0.969
LassoCV Logistic Logistic 0.220 1.180 0.974
Coverage for 95%-Confidence Interval over 1000 Repetitions
ⓘLearner g Learner m Learner r Bias CI Length Coverage
LGBM Regr. LGBM Clas. LGBM Clas. 0.247 1.112 0.932
LGBM Regr. LGBM Clas. Logistic 0.245 1.107 0.938
LGBM Regr. Logistic LGBM Clas. 0.243 1.057 0.930
LGBM Regr. Logistic Logistic 0.239 1.050 0.928
LassoCV LGBM Clas. LGBM Clas. 0.232 1.051 0.933
LassoCV LGBM Clas. Logistic 0.229 1.046 0.940
LassoCV Logistic LGBM Clas. 0.224 0.997 0.934
LassoCV Logistic Logistic 0.220 0.990 0.933
Coverage for 90%-Confidence Interval over 1000 Repetitions