README.Rmd

---
output: github_document
---

<!-- README.md is generated from README.Rmd. Please edit that file -->

```{r, include = FALSE}
knitr::opts_chunk$set(
  collapse = TRUE,
  comment = "",
  fig.path = "man/figures/README-",
  out.width = "100%",
  fig.width = 7,
  fig.height = 5,
  dev = "png",
  dpi = 600
)
```

# getspanel

<!-- badges: start -->
[![R build status](https://github.com/moritzpschwarz/getspanel/workflows/R-CMD-check/badge.svg)](https://github.com/moritzpschwarz/getspanel/actions)
[![CRAN status](https://www.r-pkg.org/badges/version/getspanel)](https://CRAN.R-project.org/package=getspanel)
<!-- badges: end -->

The package is a panel adaptation of the gets package [see here](https://CRAN.R-project.org/package=gets).

This code is being developed by Felix Pretis and Moritz Schwarz. The associated working paper is published under "Panel Break Detection: Detecting Unknown Treatment, Stability, Heterogeneity, and Outliers" by Pretis and Schwarz, which is available at SSRN [here](https://papers.ssrn.com/sol3/papers.cfm?abstract_id=4022745) and was applied to a study by Nico Koch and colleagues on EU Road CO~2~ emissions, which was [published in Nature Energy in 2022](https://www.nature.com/articles/s41560-022-01095-6). 

## Installation

You can install the released version of getspanel from [CRAN](https://cran.r-project.org/package=getspanel) with:

``` r
install.packages("getspanel")
```

And the development version from [GitHub](https://github.com/) with:

``` r
# install.packages("devtools")
devtools::install_github("moritzpschwarz/getspanel")
```


## Example

```{r example}
library(getspanel)

data("EU_emissions_road")
# let's subset a few countries to make this faster
subset <- c("Austria", "Belgium", "Germany", "Denmark", "Spain", 
                 "France",  "Greece", "Ireland", "Italy", "Netherlands", "Sweden", "United Kingdom")

EU_emissions_road <- EU_emissions_road[EU_emissions_road$country %in% subset, ]

is1 <- isatpanel(data = EU_emissions_road,
                 formula = transport.emissions ~ lgdp + lpop, 
                 index = c("country","year"),
                 effect = "twoways",
                 fesis = TRUE,
                 
                 print.searchinfo = FALSE # to save space we suppress the status information in the estimation
)
```


```{r example_textoutput}
is1
```


```{r example_plot}
plot(is1)
```

Let's explore the other plots that we can use: 

```{r}
plot_grid(is1)
```

We can plot the counterfactuals as well:

```{r}
plot_counterfactual(is1, plus_t = 5)
```

We can plot the residuals against an OLS model:

```{r}
plot_residuals(is1)
```


An example using coefficient step indicator saturation and impulse indicator saturation:

```{r, results='hide'}
is2 <- isatpanel(data = EU_emissions_road,
                 formula = transport.emissions ~ lgdp + lpop, 
                 index = c("country","year"),
                 effect = "twoways",
                 csis = TRUE, 
                 iis = TRUE,  
                 print.searchinfo = FALSE # to save space we suppress the status information in the estimation
)
```

```{r}
is2     
plot(is2)
```


```{r}
plot_grid(is2)
```


and an example of Coefficient Fixed-Effect Step indicator saturation: 

```{r, results='hide'}
is3 <- isatpanel(data = EU_emissions_road,
                 formula = transport.emissions ~ lgdp + lpop, 
                 index = c("country","year"),
                 effect = "twoways",
                 cfesis = TRUE, 
                 print.searchinfo = FALSE # to save space we suppress the status information in the estimation
)
```


```{r, results='hide'}
is3    
plot(is3)

```


We can also use e.g. the `fixest` package to estimate our models: 

```{r, results='hide'}
is4 <- isatpanel(data = EU_emissions_road,
                 formula = transport.emissions ~ lgdp + lpop, 
                 index = c("country","year"),
                 effect = "twoways",
                 engine = "fixest",
                 fesis = TRUE, 
                 print.searchinfo = FALSE # to save space we suppress the status information in the estimation
)
```


```{r, results='hide'}
plot(is4)
```