ents

Overview

This package implements methods for maximum entropy synthetic controls using multiple outcomes.

Installation

To install this package, first ensure that devtools is installed with

install.packages("devtools")

then install the package from GitHub with

devtools::install_github("ebenmichael/ents")

Basic usage

The following requires a data frame outcomes which at a minimum has the following columns:

• outcome_id: ID or name of the outcome
• time: Time that outcome was measured
• outcome: Value of the outcome
• treated: Whether the unit is treated

and a data frame metadata which at a minimum has the following columns:

• t_int: Time of intervention/treatment

Four methods of imputing a synthetic control are implemented:

• get_synth: The Abadie, Diamond, Hainmueller (2010) synthetic controls estimator, fit using Synth
• get_l2_entropy: The maximum entropy synthetic controls estimator
• get_dr: The maximum entropy synthetic controls estimator augmented with a linear outcome model
• get_ipw: An IPW estimator fit with regularized logistic regression

Each function takes in outcomes and metadata along with hyper-parameters, and returns as list including a dataframe with the synthetic control added and the synthetic control weights.