SynthControl.jl
SynthControl.jl copied to clipboard
Synthetic control methods in Julia
This pull request changes the compat entry for the `CSV` package from `0.9` to `0.9, 0.10`. This keeps the compat entries for earlier versions. Note: I have not tested your...
Registration
@JuliaRegistrator register
Described in [Ben-Michael et al. (2021)](https://www.nber.org/system/files/working_papers/w28885/w28885.pdf)
Described in [Ben-Michael et al. (2021)](https://www.nber.org/system/files/working_papers/w28885/w28885.pdf)
This is the "original" version of synthetic controls, as described in [Abadie, Diamond, Hainmueller (2010)](https://economics.mit.edu/files/11859) This should be something like `ADHSCM(x::{SingleUnitTreatment, ContinuousTreatment})` at least initially, as the paper only deals...
This is the method described in [Abadie/L'Hour (2021)](https://economics.mit.edu/files/18642), code for which is available at https://github.com/jeremylhour/pensynth It deals with multiple treated units, so should be something like `PenalizedSCM(x::BalancedPanel{Union{MultiUnitSimultaneousTreatment, MultiUnitStaggeredTreatment}, T2} where...
Initial tests seem to indicate that using JuMP to solve the convex optimization problem can yield large speedups.
Estimands: - [x] Average Treatment Effect on the Treated (ATET): ATET_t := E[Y_it(1) - Y_it(0) | i treated] - [ ] Quantile Treatment Effect on the Treated (QTET) QTET_t,alpha :=...
I coded the classic Diff in Diffs (two way Fixed Effect) outcome model, which is a special case of Synthetic Controls. I just discovered @junyuan-chen's new package [DiffinDiffs](https://github.com/JuliaDiffinDiffs/DiffinDiffs.jl).jl which has...
**Placebo methods for SE** (use MSPE of Placebo units/time to estimate SE) (3 types from [Douchenko, Imbens](https://arxiv.org/pdf/1610.07748.pdf)): - [x] Placebo unit (Random Assignment of the Unit) - [x] Placebo time...