---
title: "A bit of advanced material on R"
author: "K Foster"
date: "July 13, 2016"
output: html_document
---

This gives a bit of a taste for some of the additional features of R.


# Time Series in R
In this course we will do very little time series analysis but for now I will demonstrate some of the stuff that R can easily do.  

First install some packages (your method might vary): 
```{r eval=FALSE}
install.packages("zoo")
install.packages("lattice")
install.packages("latticeExtra")
install.packages("gdata")
install.packages("quantmod")
```

```{r message=FALSE}
library(zoo)
library(lattice)
library(latticeExtra)
library(gdata)
library(quantmod)
rm(list = ls(all = TRUE))

```

Then get data from online - also a cool easy thing to do with R.

Start with just getting data from FRED, which is easy with the quantmod package.
```{r message=FALSE}
getSymbols(c('GDP','PCEC','GPDI','GCE','NETEXP'),src='FRED')
```

Alt can load it from other government data,
```{r}
oilspot_url <- "http://www.eia.gov/dnav/pet/xls/PET_PRI_SPT_S1_D.xls"
oilspot_dat <- read.xls(oilspot_url, sheet = 2, pattern = "Cushing")

oilfut_url <- "http://www.eia.gov/dnav/pet/xls/PET_PRI_FUT_S1_D.xls"
oilfut_dat <- read.xls(oilfut_url, sheet = 2, pattern = "Cushing, OK Crude Oil Future Contract 1")
```
Use R's built-in system for converting jumbles of letters and numbers into dates,
```{r}
date_spot <- as.Date(oilspot_dat$Date, format='%b %d,%Y')
date_fut <- as.Date(oilfut_dat$Date, format='%b %d,%Y')
```
Then R's "ts" for time-series and the "zoo" package.
```{r}
wti_spot <- ts(oilspot_dat$Cushing..OK.WTI.Spot.Price.FOB..Dollars.per.Barrel., start = c(1986,2), frequency = 365)
wti_fut1 <- ts(oilfut_dat$Cushing..OK.Crude.Oil.Future.Contract.1..Dollars.per.Barrel., start = c(1983,89), frequency = 365)

wti_sp_dat <- zoo(wti_spot,date_spot)
wti_ft_dat <- zoo(wti_fut1,date_fut)

wti_spotfut <- merge(wti_sp_dat,wti_ft_dat, all=FALSE)
```
And plot the results,
```{r}
plot(wti_spotfut, plot.type = "single", col = c("black", "blue"))
```
It's tough to see any difference, so try this
```{r}
wti_2013 <- window(wti_spotfut, start = as.Date("2013-01-01"), end = as.Date("2013-12-31"))
plot(wti_2013, plot.type = "single", col = c("black", "blue"))
```
If you like this publication, you can get fancier,
```{r}
asTheEconomist(xyplot(wti_2013, xlab="Cushing WTI Spot Future Price"))
```