Skip to content

Double Machine Learning of Set-Identified Linear Models

Notifications You must be signed in to change notification settings

vsemenova/doubleml4bounds

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

24 Commits
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

doubleml4bounds

This is replication code for Semenova (2017). It estimates the upper bound on a treatment effect parameter when the outcome variable is recorded in brackets. The code compares two methods:

-- series-based support function estimator of Chandrasekhar et al (2012)

-- lasso-based double machine learning estimator of Semenova (2017)

Packages and Dependencies

The following packages must be installed:

R 3.6.1 (code was last run with version 3.6.1)

– ```hdm``` (as of 2021-02-16)
– ```xtable``` (as of 2021-02-16)

Replication of simulations

module load R

cd simulations 

Rscript Run_series.R Run_lasso.R Run_oracle_std.R Run_oracle_prop.R

Rscript print_main_table.R

Replication of empirical exercise

module load R

cd empirical_applications 

Rscript Run_lasso.R Run_forest.R

Rscript print_main_table.R

Support

Vira Semenova: semenovavira@gmail.com

About

Double Machine Learning of Set-Identified Linear Models

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

 
 
 

Languages