In this exercise you will learn to plot data using the ggplot2
package. To answer the questions below, use 4.1 Categorical vs. Categorical from Data Visualization with R.
# Load packages
library(tidyquant)
library(tidyverse)
library(lubridate) #for year()
# Pick stocks
stocks <- c("AAPL", "MSFT", "IBM")
# Import stock prices
stock_prices <- stocks %>%
tq_get(get = "stock.prices",
from = "1990-01-01",
to = "2019-05-31") %>%
group_by(symbol)
stock_prices
## # A tibble: 22,230 x 8
## # Groups: symbol [3]
## symbol date open high low close volume adjusted
## <chr> <date> <dbl> <dbl> <dbl> <dbl> <dbl> <dbl>
## 1 AAPL 1990-01-02 1.26 1.34 1.25 1.33 45799600 1.08
## 2 AAPL 1990-01-03 1.36 1.36 1.34 1.34 51998800 1.09
## 3 AAPL 1990-01-04 1.37 1.38 1.33 1.34 55378400 1.10
## 4 AAPL 1990-01-05 1.35 1.37 1.32 1.35 30828000 1.10
## 5 AAPL 1990-01-08 1.34 1.36 1.32 1.36 25393200 1.11
## 6 AAPL 1990-01-09 1.36 1.36 1.32 1.34 21534800 1.10
## 7 AAPL 1990-01-10 1.34 1.34 1.28 1.29 49929600 1.05
## 8 AAPL 1990-01-11 1.29 1.29 1.23 1.23 52763200 1.00
## 9 AAPL 1990-01-12 1.22 1.24 1.21 1.23 42974400 1.00
## 10 AAPL 1990-01-15 1.23 1.28 1.22 1.22 40434800 0.997
## # … with 22,220 more rows
# Process stock_prices and save it under stock_returns
stock_returns <-
stock_prices %>%
# Calculate yearly returns
tq_transmute(select = adjusted, mutate_fun = periodReturn, period = "yearly") %>%
# create a new variable, year
mutate(year = year(date)) %>%
# drop date
select(-date)
stock_returns
## # A tibble: 90 x 3
## # Groups: symbol [3]
## symbol yearly.returns year
## <chr> <dbl> <dbl>
## 1 AAPL 0.169 1990
## 2 AAPL 0.323 1991
## 3 AAPL 0.0691 1992
## 4 AAPL -0.504 1993
## 5 AAPL 0.352 1994
## 6 AAPL -0.173 1995
## 7 AAPL -0.345 1996
## 8 AAPL -0.371 1997
## 9 AAPL 2.12 1998
## 10 AAPL 1.51 1999
## # … with 80 more rows
Hint: See the code in 4.3.1 Bar chart (on summary statistics).
library(dplyr)
plotdata <- stock_returns %>%
group_by(symbol) %>%
summarize(mean_returns = mean(yearly.returns))
plotdata
## # A tibble: 3 x 2
## symbol mean_returns
## <chr> <dbl>
## 1 AAPL 0.366
## 2 IBM 0.116
## 3 MSFT 0.283
Hint: See the code in 4.3.1 Bar chart (on summary statistics).
ggplot(plotdata,
aes(x = symbol,
y = mean_returns)) +
geom_bar(stat = "identity")
library(scales)
ggplot(plotdata,
aes(x = factor(symbol,
labels = c("Apple",
"Microsoft",
"IBM")),
y = mean_returns)) +
geom_bar(stat = "identity",
fill = "cornflowerblue")+
geom_text(aes(label = percent(mean_returns)),
vjust = -0.25)
ggplot(stock_returns,
aes(x = yearly.returns,
fill = symbol)) +
geom_density(alpha = 0.4) +
labs(title = "Mean Yearly Returns of Apple, Microsoft and IBM")
Based on the kernel density plot, Apple has the widest distribution and lowest density, therefore it has the highest chance of losing big. Even though Microsoft and IBM have the highest density, they have less mean yearly returns than Apple, meaning their shareholders would not lose as much money if things went wrong.
Hint: See the code in 4.3.3 Box plots.
ggplot(stock_returns,
aes(x = symbol,
y = yearly.returns)) +
geom_boxplot() +
labs(title = "Distribution of Yearly Returns of Apple, Microsoft, and IBM")
If I were a risk-loving investor, I would invest in Apple because it has the highest yearly returns of the three stocks.