---
title: "Features"
author: "Eric A. Suess"
date: "2/15/2021"
output: html_document
---

```{r setup, include=FALSE}
knitr::opts_chunk$set(echo = TRUE)
```

## Time Series Features

Statistics computed on time series data are called time series features.  

```{r cars}
library(pacman)
p_load(tidyverse, fpp3)
```

We will examine the time series data in the Tourism dateset.

```{r}
data(tourism)
tourism

tourism %>% features(Trips, quantile)
```

Note the list can be used for many statistics.

```{r}
tourism %>% features(Trips, list(mean = mean, median = median, sd = sd, min = min, max = max))
```

Lets look at the first year of the data.

```{r}
tourism %>% filter(Quarter <= yearquarter("1998 Q4")) %>% 
  features(Trips, quantile)
  
```

## ACF features

The correlations in time are all time series features.

```{r}
tourism %>% ACF()
```

```{r}
tourism %>% filter(Region == "Adelaide", State == "South Australia", Purpose == "Holiday") %>% 
  ACF() 

tourism %>% filter(Region == "Adelaide", State == "South Australia", Purpose == "Holiday") %>% 
  ACF() %>% 
  autoplot()

tourism %>% filter(Region == "Adelaide", State == "South Australia", Purpose == "Holiday") %>% 
  autoplot() 
```

## Compute most of the features.

```{r}
tourism %>%
  features(Trips, feat_stl)
```

```{r}
tourism %>% filter(Region == "Adelaide", State == "South Australia", Purpose == "Holiday") %>% 
  features(Trips, feat_stl) 
```


## Use the features to identify a time series

We can then use these features in plots to identify what type of series are heavily trended and what are most seasonal.

```{r}
tourism %>%
  features(Trips, feat_stl) %>%
  ggplot(aes(x = trend_strength, y = seasonal_strength_year, col = Purpose)) +
  geom_point() +
  facet_wrap(vars(State))
```

Find the year with the maximum seasonal strength.

```{r}
tourism %>%
  features(Trips, feat_stl) %>%
  filter(seasonal_strength_year == max(seasonal_strength_year)) %>%
  left_join(tourism, by = c("State", "Region", "Purpose")) %>%
  ggplot(aes(x = Quarter, y = Trips)) +
  geom_line() +
  facet_grid(vars(State, Region, Purpose))
```

## Full feature set

```{r}
tourism_features <- tourism %>%
  features(Trips, feature_set(pkgs = "feasts"))
tourism_features
```

For the homework we will look at the PBS data.

```{r}
PBS %>% features(Cost, list(mean = mean, sd = sd))
```