Multiple Lines on Multiple Plots in R

Introduction

In this article, we will explore how to create multiple plots with different time series data and a shared x-axis. We will discuss the use of ggplot2 and some creative solutions to achieve this.

Background

When working with multiple time series data, it can be challenging to visualize them in a single plot without overwhelming the viewer. One common approach is to create separate plots for each time series using libraries like par(mfrow) or facet_wrap(). However, when we want to display multiple lines on the same plot with different time series data, things become more complicated.

In this article, we will delve into the world of ggplot2 and explore ways to create multiple plots with shared x-axes using creative solutions like faceting and grid arrangement.

Creating Multiple Plots with Shared X-Axis

Let’s start by examining a basic example using par(mfrow). We’ll generate four random time series data and plot them using plot.ts().

# Load required libraries
library(TTR)
library(ggplot2)

# Generate random time series data
ts1 <- rnorm(100)
ts2 <- rnorm(100)
ts3 <- rnorm(100)
ts4 <- rnorm(100)

# Create a dataframe with the data and IDs
df <- data.frame(time = rep(1:100, 8), 
                 id = as.factor(rep(1:8, each = 100)), 
                 type = as.factor(rep(1:2, each = 200)),
                 value = c(ts1, SMA(ts1), ts2, SMA(ts2), ts3, SMA(ts3), ts4, SMA(ts4)))

# Create the plot
par(mfrow = c(4, 1))
plot.ts(ts1, col = "green")
lines(SMA(ts1, n = 10), col = "red")
plot.ts(ts2, col = "green")
lines(SMA(ts2, n = 10), col = "red")
plot.ts(ts3, col = "green")
lines(SMA(ts3, n = 10), col = "red")
plot.ts(ts4, col = "green")
lines(SMA(ts4, n = 10), col = "red")

As you can see, the plot has multiple lines but also some awkward labels on the x-axis. This is because we’re using par(mfrow) to create multiple plots on the same graph.

Using ggplot2

Now, let’s move on to using ggplot2. We’ll generate the same random time series data and create a dataframe with it.

# Load required libraries
library(ggplot2)

# Generate random time series data
ts1 <- rnorm(100)
ts2 <- rnorm(100)
ts3 <- rnorm(100)
ts4 <- rnorm(100)

# Create a dataframe with the data and IDs
df <- data.frame(time = rep(1:100, 8), 
                 id = as.factor(rep(1:8, each = 100)), 
                 type = as.factor(rep(1:2, each = 200)),
                 value = c(ts1, SMA(ts1), ts2, SMA(ts2), ts3, SMA(ts3), ts4, SMA(ts4)))

# Create the plot
ggplot(df, aes(time, value, col = type, group = id)) + 
  geom_line() + 
  facet_wrap(~id1, ncol = 1) + 
  scale_color_manual(values = c('green', 'red')) +
  guides(color = FALSE) + 
  theme_bw() + 
  theme(strip.text = element_blank())

As you can see, the plot has multiple lines with shared x-axes. This is achieved using facet_wrap() to create separate plots for each time series ID.

One potential limitation of ggplot2 is that it doesn’t allow us to add labels to the x-axis directly when faceting. However, we can use the grid.arrange() function from the gridExtra package to create a custom layout.

# Load required libraries
library(ggplot2)
library(gridExtra)

# Create the plot
ggplot(data.frame(time = rep(1:100, 2), 
                 id = as.factor(rep(1:2, each = 100)), 
                 type = as.factor(rep(1:2, each = 100)), 
                 ts1 = c(ts1, SMA(ts1))), aes(time, ts1, col = type, group = id)) + 
  geom_line() + 
  scale_color_manual(values = c('green', 'red')) +
  guides(color = FALSE) +
  theme_bw() + 
  theme(axis.text.x = element_blank(), axis.ticks = element_blank()) + xlab('') +
  theme(axis.text.x = element_blank(), axis.ticks = element_blank())

ggplot(data.frame(time = rep(1:100, 2), 
                 id = as.factor(rep(1:2, each = 100)), 
                 type = as.factor(rep(1:2, each = 100)), 
                 ts2 = c(ts2, SMA(ts2))), aes(time, ts2, col = type, group = id)) + 
  geom_line() + 
  scale_color_manual(values = c('green', 'red')) +
  guides(color = FALSE) +
  theme_bw() + 
  theme(axis.text.x = element_blank(), axis.ticks = element_blank()) + xlab('') +
  theme(axis.text.x = element_blank(), axis.ticks = element_blank())

ggplot(data.frame(time = rep(1:100, 2), 
                 id = as.factor(rep(1:2, each = 100)), 
                 type = as.factor(rep(1:2, each = 100)), 
                 ts3 = c(ts3, SMA(ts3))), aes(time, ts3, col = type, group = id)) + 
  geom_line() + 
  scale_color_manual(values = c('green', 'red')) +
  guides(color = FALSE) +
  theme_bw() + 
  theme(axis.text.x = element_blank(), axis.ticks = element_blank()) + xlab('') +
  theme(axis.text.x = element_blank(), axis.ticks = element_blank())

ggplot(data.frame(time = rep(1:100, 2), 
                 id = as.factor(rep(1:2, each = 100)), 
                 type = as.factor(rep(1:2, each = 100)), 
                 ts4 = c(ts4, SMA(ts4))), aes(time, ts4, col = type, group = id)) + 
  geom_line() + 
  scale_color_manual(values = c('green', 'red')) +
  guides(color = FALSE) +
  theme_bw()

As you can see, the plot has multiple lines with shared x-axes and no labels on the x-axis. This is achieved using grid.arrange() to create a custom layout.

Conclusion

In this article, we explored how to create multiple plots with different time series data and a shared x-axis in R. We discussed the use of ggplot2 and some creative solutions like faceting and grid arrangement to achieve this. Whether you’re working with large datasets or just want to visualize your data effectively, these techniques will help you create informative and engaging plots.

Last modified on 2024-09-02