---
title: "Introduction to the tsdl package"
author: "Yangzhuoran Yang"
date: "`r Sys.Date()`"
output: rmarkdown::html_vignette
vignette: >
  %\VignetteIndexEntry{Introduction to the tsdl package}
  %\VignetteEngine{knitr::rmarkdown}
  %\VignetteEncoding{UTF-8}
---

```{r setup, include = FALSE}
knitr::opts_chunk$set(
  collapse = TRUE,
  comment = "#>", 
  fig.align = "center"
)
library(knitr)
```


The Time Series Data Library (TSDL) was created by Rob Hyndman, Professor of Statistics at Monash University, Australia. It includes data from a lot of time series textbooks, as well as many other series that he has either collected for student projects or helpful people have sent to him.  

The data library was once hosted on Professor Hyndman's personal website since about 1992. In 2012 [it was moved](https://robjhyndman.com/hyndsight/tsdl/) onto [DataMarket](https://datamarket.com/data/list/?q=provider:tsdl) which provides much better facilities for maintaining and using time series data. You can still access the data library from the website or using `rdatamarket` package to read it into R, but `tsdl` package provides a simpler means.  

If you use any data from the TSDL in a publication, please use the following citation:  
 
> Rob Hyndman and Yangzhuoran Yang (2018). tsdl: Time Series Data Library. v0.1.0. https://pkg.yangzhuoranyang./tsdl/.  

The data files will remain on Professor Hyndman's personal website so that existing links will not be broken.  

## Installation


You can install the **development** version from
[Github](https://github.com/FinYang/tsdl)

```r
# install.packages("devtools")
devtools::install_github("FinYang/tsdl")
```

## Usage

`tsdl` is a list of 648 series of class `tsdl`. Each series within tsdl is of class `ts`.

```{r usage}
library(tsdl)
tsdl
```

To extract series with specific features, one can use function `subset`. The most common way to extract series is to specify `frequency` or `subject` (type) of the series. The position of these two set conditions are interchangeable.  

```{r}
# Subset by frequency
tsdl_quarterly <- subset(tsdl,4)
tsdl_quarterly

# Subset by subject
tsdl_industry <- subset(tsdl,"Industry")
tsdl_industry

# Subset by frequency and subject
tsdl_daily_industry <- subset(tsdl,12,"Industry")
tsdl_daily_industry
```

User can also subset the data set using specified `start` year, or keywords in its `source` attribute or `description` attribute.  

```{r}
# Subset by source
tsdl_abs <- subset(tsdl, source = "Australian Bureau of Statistics")
tsdl_abs

# Subset by starting year
tsdl_1948 <- subset(tsdl, start = 1948)
tsdl_1948

# Subset by description
tsdl_nettraffic <- subset(tsdl, description = "Internet traffic")
tsdl_nettraffic
```

To access attributes information of the time series, one can directly extract its attributes.

```{r}
attributes(tsdl[[1]])
```

The collective attributes information is stored in the data frame `meta_tsdl`. One can also access the possible choices of `subject` and other options when subset time series.  

```{r}
str(meta_tsdl)
unique(meta_tsdl$subject)
```




## License

This package is free and open source software, licensed under GPL-3