-
Notifications
You must be signed in to change notification settings - Fork 0
/
Copy pathREADME.rmd
62 lines (39 loc) · 3.64 KB
/
README.rmd
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
---
output: github_document
---
# Repository for the paper "Survey calibration for causal inference: a simple method to balance covariate distributions"
# Paper
Here is the first version of the paper: ["Survey calibration for causal inference: a simple method to balance covariate distributions"](paper/2023-beresewicz-causal-balancing-new.pdf) (2023.10.14, [arxiv version](https://arxiv.org/abs/2310.11969))
# Acknowledgements
This work was financed by the National Science Centre in Poland, OPUS 22 grant no. 2020/39/B/HS4/00941.
# Reproduction of the results
## Setup
Install relevant packages for the paper. The `jointCalib` package is available at CRAN but we install the package from github (development version). The `IPS` package used in the [Sant'Anna et al. (2022)](https://onlinelibrary.wiley.com/doi/10.1002/jae.2909) and the `kbal` package used in the [Hazlett (2020)](https://www3.stat.sinica.edu.tw/statistica/j30n3/J30N32/J30N32.html) are available only on github.
**UPDATE**: as of 2023.10.19 (version: 0.14.2.9001) the `WeightIt` package implements methods proposed in my paper (yay!). See [WeightIt NEWS.md file](https://github.com/ngreifer/WeightIt/blob/master/NEWS.md) and the documentation of `ebal`, `npcbps`, `optweight` and `energy`. Importantly, [Noah Greifer](https://github.com/ngreifer) implementation allows for multinomial treatment studies.
```{r install-packages, eval = FALSE}
install.packages(c("remotes", "ebal", "mvnfast", "data.table", "ggplot2", "laeken", "xtable", "glue", "stringr"))
remotes::install_github("ncn-foreigners/jointCalib@dev")
remotes::install_github("chadhazlett/KBAL")
remotes::install_github("pedrohcgs/IPS")
remotes::install_github("ngreifer/WeightIt") ## 0.14.2.9001
```
## Notebooks and results
+ Notebooks:
+ [simulation 1 (EB)](https://htmlpreview.github.io/?https://raw.githubusercontent.com/ncn-foreigners/paper-note-quantiles-obs-studies/main/codes/1-simulation-eb.html),
+ [simulation 2 (CBPS)](https://htmlpreview.github.io/?https://raw.githubusercontent.com/ncn-foreigners/paper-note-quantiles-obs-studies/main/codes/2-simulation-ps.html)
+ [processing results for the paper](https://htmlpreview.github.io/?https://raw.githubusercontent.com/ncn-foreigners/paper-note-quantiles-obs-studies/main/codes/3-report-results-mc.html)
+ [simulation 3 (why it works)](https://htmlpreview.github.io/?https://raw.githubusercontent.com/ncn-foreigners/paper-note-quantiles-obs-studies/main/codes/4-why-it-works.html)
+ Simulation results may be found in folder `results/`
+ Tutorial for [the `jointCalib` package](https://ncn-foreigners.github.io/jointCalib/articles/d_causal.html)
+ A minimal example using the [`WeightIt`](https://htmlpreview.github.io/?https://raw.githubusercontent.com/ncn-foreigners/paper-note-quantiles-obs-studies/main/codes/6-weightit.html) package.
+ Codes for the proposed method in [`Stata`](https://htmlpreview.github.io/?https://raw.githubusercontent.com/ncn-foreigners/paper-note-quantiles-obs-studies/main/codes/5-minimal-code-stata.html) and `Python` [work in progress].
## Highlights
### Results for the distributional entropy balancing (DEB)
### Results for the distributional propensity score (DPS)
### Why it works?
Including quartiles ($A_q$) and deciles ($A_d$) along with X approximates the relationship between Y and X by local linear functions (as in segmented regression) and thus improves the estimates, as EB and CBPS assume a linear relationship between Y and calibration/balance variables.
