tutorial-long-run-filters.Rmd

---
title: "Long Run vs Short Run Decompositions in R"
author: "Carlos Mendez"
subtitle: The HP filter vs the Hamilton filter
output:
  html_document:
    code_folding: show
    highlight: monochrome
    number_sections: yes
    theme: cosmo
    toc: yes
    toc_depth: 4
    toc_float:
      collapsed: no
      smooth_scroll: no
  github_document: default
always_allow_html: true
---

<style>
h1.title {font-size: 18pt; color: DarkBlue;} 
body, h1, h2, h3, h4 {font-family: "Palatino", serif;}
body {font-size: 12pt;}
/* Headers */
h1,h2,h3,h4,h5,h6{font-size: 14pt; color: #00008B;}
body {color: #333333;}
a, a:hover {color: #8B3A62;}
pre {font-size: 12px;}
</style>


Suggested Citation: 

> Mendez C. (2020). Long Run vs Short Run Decompositions in R: The HP filter vs the Hamilton filter. R Studio/RPubs. Available at <https://rpubs.com/quarcs-lab/long-run-filters>

This work is licensed under the Creative Commons Attribution-Share Alike 4.0 International License. 
![](License.png)




# Set parameters of the program

- Name of the series

```{r}
seriesName <- "RGDPNAIDA666NRUG"
```

Code examples for other series

- Total GDP of Japan: "JPNRGDPEXP"
- GDP per capita of Japan: "RGDPCHJPA625NUPN"
- GPD per capita of Bolivia:  "NYGDPPCAPKDBOL"
- Total GDP of Bolivia: "RGDPNABOA666NRUG"
- Total GDP of Indonesia: "RGDPNAIDA666NRUG"



# Load libraries

```{r message=FALSE, warning=FALSE}
library(mFilter)
library(quantmod)
library(dplyr)
library(ggplot2)
library(dygraphs)
library(xts)
library(neverhpfilter)

# Change the presentation of decimal numbers to 4 and avoid scientific notation
options(prompt="R> ", digits=3, scipen=999)
```


# Import data

```{r}
seriesName <- getSymbols(seriesName, src="FRED", auto.assign = FALSE)
```

```{r}
periodicity(seriesName)
```



# Transform the data

- Take the log of the series

```{r}
seriesName <- log(seriesName)
```


# Plot evolution of the variable


```{r}
dygraph(seriesName) %>% 
  dyRangeSelector()
```



# Apply the HP filter


```{r}
seriesName_filtered_HP <- hpfilter(seriesName, 
                                  freq = 6.25      
                                     )
```

## Plot the HP filter

### Long-run trend

Create matrix of actual, trend , and cycle values

```{r}
actual  <- seriesName_filtered_HP[["x"]]
trendHP <- seriesName_filtered_HP[["trend"]]
cycleHP <- actual - trendHP

colnames(actual)   <- c("actual")
colnames(trendHP)  <- c("trendHP")
colnames(cycleHP)  <- c("cycleHP")

actual_and_trend <- cbind(actual, trendHP)
```


```{r}
dygraph(actual_and_trend[,1:2]) %>%
  dyRangeSelector()
```




### Short-run fluctuations

```{r}
dygraph(cycleHP)  %>% 
  dyRangeSelector()
```


# Apply the Hamilton filter

```{r}
seriesName_filtered_Hamilton <- yth_filter(seriesName, 
                                           h = 2, 
                                           p = 4, 
                                           output = c("x", "trend", "cycle"))
```


## Plot the Hamiltion filter

### Long-run trend

Rename columns

```{r}
colnames(seriesName_filtered_Hamilton)  <- c("actual",
                                             "trendHamilton",
                                             "cycleHamilton")
```


```{r}
dygraph(seriesName_filtered_Hamilton[,1:2])  %>% 
  dyRangeSelector()
```


### Short-run fluctuation

```{r}
dygraph(seriesName_filtered_Hamilton[,3])  %>% 
  dyRangeSelector()
```



# Run it in the cloud

Tip: Copy and paste this link another tab of your browser.

<https://rstudio.cloud/project/25043>

Or simply <a href="https://rstudio.cloud/project/25043" target="_blank">ClickHERE</a>

# References 

- [Schüler, Y., 2018. On the cyclical properties of Hamilton's regression filter (No. 03/2018). Deutsche Bundesbank.](https://www.econstor.eu/bitstream/10419/174891/1/1014338883.pdf)

- <http://past.rinfinance.com/agenda/2018/JustinShea.pdf>

- <https://www.r-econometrics.com/timeseries/economic-cycle-extraction/>

- <https://www.datacamp.com/community/blog/r-xts-cheat-sheet>

- <https://rstudio.github.io/dygraphs/>