R Notebook

R Time series examples.

1) Apple Stock Price

• Start by importing the proper libraries and loading in some data to start with. I start by pulling Apple stock prices from 1990 up to today. Stock market data is fun and easy to play with.

# Load libraries
library(sweep)      # Broom-style tidiers for the forecast package
library(forecast)   # Forecasting models and predictions package
library(tidyquant)  # Loads tidyverse, financial pkgs, used to get data
library(timetk)     # Functions working with time series
library(tidyverse)  # Functions for data analysis
AAPL = tq_get("AAPL", get = "stock.prices", from = " 1990-01-01")
head(AAPL)

Apple Closing from 1990 up to today. In general it looks like there is an upward trend in price.

AAPL %>% 
  ggplot(aes(date, adjusted))+
  geom_line()+
  labs(title = "Apple Adjusted Closing Stock Price", x = "Date", y = "APPL   Close Price") + 
  theme_economist() + 
  stat_smooth(method = "loess", alpha = 0.2)

I next look at the trading volume of Apple stock. I’ll start by plotting a histogram of the closing price.

AAPL %>% 
  ggplot(aes(volume)) +
  geom_histogram(bins = 100)+
  geom_vline(xintercept = mean(AAPL$volume), color = "salmon")+
  theme_economist()+
  labs(title = "Histogram of AAPL Volume")

2) Zillow Home Value Index Vermont:

VT_home =
    "ZILLOW/S46_TURNAH" %>%
    tq_get(get  = "quandl",
    from = "2008-01-01")

VT_home %>% ggplot(aes(date, value))+
  geom_line()+
  labs(title = "Zillow Home Value Index (State):", subtitle = "Turnover -   All Homes - Vermont", x = "Year", y= "Zillow Home Value Index")+
  theme_economist()

NA

3) Skateboard Industry Stock Analysis

Stocks contained:

• Vans Footwear Company - VFC

• Sequential Brands Group / DVS - SQBG

• Nike - NKE

• Iconix / Zoo York - ICON

• Zumiez Retail - ZUMZ

symbols = c("VFC", "SQBG", "NKE", "ICON", "ZUMZ")
skate = symbols %>%  tq_get(get = "stock.prices", from = " 2002-01-01")

skate %>% ggplot(aes(date, adjusted)) +
geom_line() +
geom_point(alpha= 0.2, size=0.2)+
facet_wrap(~symbol, nrow = 5)+
theme_economist()+
scale_x_date()+
labs(title = "Skateboard Industry Adjusted Stock Prices 2002 to 2018")

skate %>% ggplot(aes(date, volume)) +
geom_line(color = "black") +
facet_wrap(~symbol, nrow = 5)+
theme_economist()+
labs(title = "Skateboard Industry Volume")

skate %>% ggplot(aes(adjusted)) +
geom_histogram(bins = 30) +
facet_wrap(~symbol, nrow = 5)+
theme_economist()+
labs(title = "Skateboard Industry Adjusted Price Histograms")

vans = skate %>% filter(date > mdy(01012012) & symbol == "VFC") %>% select(date, adjusted)
colnames(vans) = c("ds", "y")
vans$adjusted = log10(vans$y)

library(prophet)
m <- prophet(vans)

Disabling daily seasonality. Run prophet with daily.seasonality=TRUE to override this.

Initial log joint probability = -13.2182
Optimization terminated normally: 
  Convergence detected: relative gradient magnitude is below tolerance

future <- make_future_dataframe(m, periods = 365)
forecast <- predict(m, future)

|==========================================|100% ~0 s remaining     

library(plotly)
ggplotly(
plot(m, forecast) + theme_economist() + labs(title = "Vans Footwear Adjusted Stock Price Forecast", x = "Date", y = "Adjusted Stock Price"))

We recommend that you use the dev version of ggplot2 with `ggplotly()`
Install it with: `devtools::install_github('hadley/ggplot2')`

prophet_plot_components(m, forecast) 

Nike forecast

Nike = skate %>% filter(date > mdy(01012012) & symbol == "NKE") %>% select(date, adjusted)
colnames(Nike) = c("ds", "y")
vans$adjusted = log10(vans$y)

library(prophet)
m2 <- prophet(Nike)

Disabling daily seasonality. Run prophet with daily.seasonality=TRUE to override this.

Initial log joint probability = -6.39235
Optimization terminated normally: 
  Convergence detected: relative gradient magnitude is below tolerance

future2 <- make_future_dataframe(m2, periods = 365)
forecast2 <- predict(m2, future)

|==========================================|100% ~0 s remaining     

plot(m2, forecast2) + theme_economist() + labs(title = "Nike Adjusted Stock Price Forecast", x = "Date", y = "Adjusted Stock Price")

prophet_plot_components(m2, forecast2) 

library(sparklyr)