Visualization Best Practices in R
Beginner
4 hr
14.3K learners
In this tutorial, we will be visualizing distributions of data by plotting histograms using the R programming language. We will cover what a histogram is, how to read data in R, how to create a histogram, and how to customize the plot.
We will be using the base R programming language with no additional packages. This approach is only recommended when additional packages cannot be used or for quick exploratory analyses. In most cases, you should use ggplot2, as covered in the How to make a histogram in R in ggplot2 tutorial.
What is a histogram?
A histogram is a very popular graph that is used to show frequency distributions across continuous (numeric) variables. Histograms allow us to see the count of observations in data within ranges that the variable spans.
Histograms look similar to bar charts. A key difference between the two is that bar charts have a value associated with a specific category or discrete variable, while a histogram visualizes frequencies for continuous variables.
Housing Data
We will be using this housing dataset which includes details about different house listings, including the size of the house, the number of rooms, the price, and location information. We can read the data using the read.csv() function, either directly from the URL or by downloading the csv file into a directory and reading it from our local storage. We can also specify that we only want to store the columns we are interested in for this tutorial; price and condition.
home_data <- read.csv("https://raw.githubusercontent.com/rashida048/Datasets/master/home_data.csv")[ ,c('price', 'condition')]Let’s look at the first few rows of data using the head() function
head(home_data, 5)
How to Make a Histogram with base R using the hist() function
Next, we will create a histogram using the hist() function to look at the distribution of prices in our dataset.
hist(home_data$price)
Add descriptive statistics to histogram
We can add descriptive statistics to the histogram using the abline() function. This adds a vertical line to the plot.
- Set the v argument to the position on the x-axis for the vertical line. Here, we get the mean house price using
mean(). - The
colargument set the line color, in this case to red. - The
lwdargument sets the line width. A value of 3 increases the thickness of the line to make it easier to see.
hist(home_data$price)
abline(v = mean(home_data$price), col='red', lwd = 3)Plotting probability densities instead of counts
To add a probability density line to the histogram, we first change the y-axis to be scaled to density. In the call to hist() , we set the probability argument to TRUE.
The probability density line is made with a combination of density(), which calculates the position of the probability density curve, and lines(), which adds the line to the existing plot.
hist(home_data$price, probability = TRUE)
abline(v = mean(home_data$price), col='red', lwd = 3)
lines(density(home_data$price), col = 'green', lwd = 3)
Notice that the numbers on the y-axis have changed.
Customize the color of the histogram using col
We can change the colors inside of the bins on the histogram using the col parameter of the hist() function. We will change the fill to blue. We can also change the outline color of the bars using the border parameter. We will change the color of the outlines to white.
hist(home_data$price, col = 'blue', border = "white")
Add labels and titles using ylab, xlab, and main
We can change the labels on the plot to make it more readable and presentable. This is useful if you share the plot with others.
-
xlabsets the x-axis label -
ylabsets the y-axis label -
mainsets the plot title
hist(home_data$price, xlab = 'Price (USD)', ylab = 'Number of Listings', main = 'Distribution of House Prices')
Start Learning R For Free
Introduction to Data Visualization with ggplot2
Beginner
4 hr
105.4K learners
Update binning using breaks
With the default arguments, it is challenging to see the full distribution of the housing prices across the range of prices. We can see they are centralized in the first few bins, but they are not very descriptive.
We can add more bins using the breaks parameter. With this argument, we can pass a vector of specific breakpoints to use, a function to compute the breakpoints, a number of breaks we would like, or a function to compute the number of cells.
For this example, we will pass the number of bins we would like. This number is context-specific based on what you are trying to show in your graph.
hist(home_data$price, breaks = 100)
With breaks set to 100, we have significantly more visibility into the distribution in the first few buckets.
We can also specify the number of breaks using the names of common calculations for calculating optimal breaks in a histogram. By default, hist() uses the “Sturges” method (this is optional to pass as an argument because it is the default).
hist(home_data$price, breaks = "Sturges")
We can also pass “Scott” as an argument for the breaks attribute to use the Scott Method.
hist(home_data$price, breaks = "Scott")
Finally, we could also use the Freedman-Diaconis (FD) method.
hist(home_data$price, breaks = "Freedman-Diaconis")
Setting x-axis limits
We can set the x-axis limits of our plot using the xlim argument to zoom in on the data we are interested in. For example, it is sometimes helpful to focus on the central part of the distribution, rather than over the long tail we currently see when we view the whole plot.
Changing the y-axis limits is also possible (using the ylim argument) but this is less useful for histograms since the automatically calculated values are almost always ideal.
We will zoom in on prices between $0 and $2M.
hist(home_data$price, breaks = 100, xlim = c(0, 2000000))
Take it to the next level
As you get more comfortable with R, you can explore more powerful packages that make it easier to build more interesting and useful visualizations. A very popular and easy-to-use library for plotting in R is called ggplot2. Below we create an interesting view of the distributions of prices based on the number of bedrooms in the house.
ggplot2 is the best way to visualize data in R, and you can learn about using it to create histograms in the How to make a histogram in R in ggplot2 tutorial. Check out our Introduction to ggplot2 course and our Intermediate ggplot2 course to learn how to make more interesting visualizations in R.
In summary
In this tutorial, we learned that histograms are great visualizations for looking at distributions of continuous variables. We learned how to make a histogram in R, how to plot summary statistics on top of our histogram, how to customize features of the plot like the axis titles, the color, how we bin the x-axis, and how to set limits on the axes. Finally, we demonstrated some of the power of the ggplot2 library.
Further Reading and Resources for Histogram Plotting in R:
R Courses
Introduction to R
Beginner
4 hr
2.5M
Intermediate R
Beginner
6 hr
545.3K
Exploratory Data Analysis in R
Beginner
4 hr
83.5K
See More
RelatedSee MoreSee More
Data-Driven Thinking for Everyday Life
Gary Wolf talks about what The Quantified Self is, why self-tracking projects can be life-changing, how to get started with self-tracking, and how to connect with others.
Richie Cotton
55 min
Reshaping Data with pandas in Python
Pandas DataFrames are commonly used in Python for data analysis, with observations containing values or variables related to a single object and variables representing attributes across all observations.
Richie Cotton
Reshaping Data with tidyr in R
In this cheat sheet, you will learn how to reshape data with tidyr. From separating and combining columns, to dealing with missing data, you'll get the download on how to manipulate data in R.
Richie Cotton
6 min
Data Quality Dimensions Cheat Sheet
In this cheat sheet, you'll learn about data quality dimensions, allowing you to ensure that your data is fit for purpose.
Joe Franklin
3 min
Top Techniques to Handle Missing Values Every Data Scientist Should Know
Explore various techniques to efficiently handle missing values and their implementations in Python.
Zoumana Keita
15 min
T-tests in R Tutorial: Learn How to Conduct T-Tests
Determine if there is a significant difference between the means of the two groups using t.test() in R.
Abid Ali Awan
10 min