This is the init_notebook_mode cell from ITables v2.1.4
(you should not see this message - is your notebook trusted?)
ATTE Coverage
The simulations are based on the the make_did_SZ2020-DGP with \(1000\) observations. Learners are only set to boosting, due to time constraints (and the nonlinearity of some of the DGPs).
DoubleML Version 0.8.2
Script did_pa_atte_coverage.py
Date 2024-08-12 14:36:11
Total Runtime (seconds) 13934.719484
Python Version 3.12.4
Observational Score
Coverage for 95.0%-Confidence Interval over 1000 Repetitions
| LGBM |
LGBM |
1 |
False |
3.229 |
14.583 |
0.960 |
| LGBM |
LGBM |
2 |
False |
3.537 |
17.722 |
0.971 |
| LGBM |
LGBM |
3 |
False |
3.347 |
17.105 |
0.974 |
| LGBM |
LGBM |
4 |
False |
5.443 |
21.039 |
0.953 |
| LGBM |
LGBM |
5 |
False |
1.927 |
9.351 |
0.975 |
| LGBM |
LGBM |
6 |
False |
1.845 |
9.188 |
0.974 |
| LGBM |
LGBM |
1 |
True |
1.038 |
4.993 |
0.961 |
| LGBM |
LGBM |
2 |
True |
1.212 |
5.940 |
0.962 |
| LGBM |
LGBM |
3 |
True |
1.122 |
5.753 |
0.972 |
| LGBM |
LGBM |
4 |
True |
1.300 |
7.104 |
0.981 |
| LGBM |
LGBM |
5 |
True |
0.859 |
4.237 |
0.967 |
| LGBM |
LGBM |
6 |
True |
0.822 |
4.050 |
0.960 |
Coverage for 90.0%-Confidence Interval over 1000 Repetitions
| LGBM |
LGBM |
1 |
False |
3.229 |
12.239 |
0.911 |
| LGBM |
LGBM |
2 |
False |
3.537 |
14.873 |
0.922 |
| LGBM |
LGBM |
3 |
False |
3.347 |
14.355 |
0.938 |
| LGBM |
LGBM |
4 |
False |
5.443 |
17.657 |
0.857 |
| LGBM |
LGBM |
5 |
False |
1.927 |
7.847 |
0.916 |
| LGBM |
LGBM |
6 |
False |
1.845 |
7.710 |
0.926 |
| LGBM |
LGBM |
1 |
True |
1.038 |
4.191 |
0.908 |
| LGBM |
LGBM |
2 |
True |
1.212 |
4.985 |
0.919 |
| LGBM |
LGBM |
3 |
True |
1.122 |
4.828 |
0.924 |
| LGBM |
LGBM |
4 |
True |
1.300 |
5.962 |
0.939 |
| LGBM |
LGBM |
5 |
True |
0.859 |
3.556 |
0.911 |
| LGBM |
LGBM |
6 |
True |
0.822 |
3.399 |
0.904 |
Experimental Score
Remark that the only two valid DGPs are DGP \(5\) and DGP \(6\). All other DGPs are invalid due to non-experimental treatment assignment.
Coverage for 95.0%-Confidence Interval over 1000 Repetitions
| LGBM |
LGBM |
1 |
False |
2.175 |
2.578 |
0.087 |
| LGBM |
LGBM |
2 |
False |
1.322 |
2.540 |
0.483 |
| LGBM |
LGBM |
3 |
False |
0.999 |
2.241 |
0.591 |
| LGBM |
LGBM |
4 |
False |
1.928 |
2.248 |
0.124 |
| LGBM |
LGBM |
5 |
False |
0.522 |
2.461 |
0.942 |
| LGBM |
LGBM |
6 |
False |
0.438 |
2.155 |
0.949 |
| LGBM |
LGBM |
1 |
True |
2.169 |
2.577 |
0.089 |
| LGBM |
LGBM |
2 |
True |
1.307 |
2.534 |
0.486 |
| LGBM |
LGBM |
3 |
True |
0.999 |
2.239 |
0.606 |
| LGBM |
LGBM |
4 |
True |
1.924 |
2.246 |
0.111 |
| LGBM |
LGBM |
5 |
True |
0.523 |
2.462 |
0.949 |
| LGBM |
LGBM |
6 |
True |
0.425 |
2.156 |
0.953 |
Coverage for 90.0%-Confidence Interval over 1000 Repetitions
| LGBM |
LGBM |
1 |
False |
2.175 |
2.163 |
0.048 |
| LGBM |
LGBM |
2 |
False |
1.322 |
2.131 |
0.376 |
| LGBM |
LGBM |
3 |
False |
0.999 |
1.880 |
0.479 |
| LGBM |
LGBM |
4 |
False |
1.928 |
1.887 |
0.083 |
| LGBM |
LGBM |
5 |
False |
0.522 |
2.065 |
0.878 |
| LGBM |
LGBM |
6 |
False |
0.438 |
1.809 |
0.903 |
| LGBM |
LGBM |
1 |
True |
2.169 |
2.163 |
0.054 |
| LGBM |
LGBM |
2 |
True |
1.307 |
2.127 |
0.388 |
| LGBM |
LGBM |
3 |
True |
0.999 |
1.879 |
0.477 |
| LGBM |
LGBM |
4 |
True |
1.924 |
1.884 |
0.076 |
| LGBM |
LGBM |
5 |
True |
0.523 |
2.066 |
0.889 |
| LGBM |
LGBM |
6 |
True |
0.425 |
1.809 |
0.906 |