Preparing your figures to share with others

1. Preparing your figures to share with others

This chapter will focus on creating visualizations that you can share with others and incorporate into automated data analysis pipelines. We'll start with customization of figure styles. Previously, you saw that you can change the appearance of individual elements of the figure, such as the line color, or marker shapes.

2. Changing plot style

Here, we'll change the overall style of the figure. To see what that means, let's look at one of the figures we created in a previous lesson. This figure shows the average temperatures in Seattle and Austin as a function of the months of the year. This is what it looks like per default.

3. Choosing a style

If instead, we add this line of code before the plotting code, the figure style will look completely different. The style we chose here emulates the style of the R library ggplot. Maybe you know this library and this looks familiar to you, or you can learn about ggplot in a DataCamp course devoted to this library. Either way, you will notice that the setting of the style didn't change the appearance of just one element in the figure. Rather, it changed multiple elements: the colors are different, the fonts used in the text are different, and there is an added gray background that creates a faint white grid marking the x-axis and y-axis tick locations within the plot area. Furthermore, this style will now apply to all of the figures in this session, until you change it by choosing another style.

4. Back to the default

For example, to go back to the default style, you would run plt-dot-style-dot-use "default".

5. The available styles

Matplotlib contains implementations of several different styles and you can see the different styles available by going to this webpage, which contains a series of visualizations that have each been created using one of the available styles.

6. The "bmh" style

For example, this is what you get if you use "bmh" as the style.

7. Seaborn styles

This is what you get if you select "seaborn-colorblind". In fact, if you visit the documentation web-page, you will see that there are several available styles that are named after the Seaborn software library. This is a software library for statistical visualization that is based on Matplotlib, and Matplotlib adopted back several of the styles developed there. You can learn more about Seaborn in other DataCamp courses.

8. Guidelines for choosing plotting style

How would you choose which style to use? If your goal is primarily to communicate with others, think about how they might see it. Dark backgrounds are generally discouraged as they are less visible, so only use them if you have a good reason to do so. If colors are important, consider using a colorblind-friendly style, such as "seaborn-colorblind" or "tableau-colorblind10". These are designed to retain color differences even when viewed by colorblind individuals. That might sound like a minor consideration, but approximately 1 out of 20 individuals is colorblind. Figures that are designed for use on websites have different considerations than figures in printed reports. For example, if someone is going to print out your figures, you might want to use less ink. That is, avoid colored backgrounds, like the background that appears in the "ggplot" style that we demonstrated before. If the printer used is likely to be black-and-white, consider using the "grayscale" style. This will retain the differences you see on your screen when printed out in a black-and-white printer.

9. Practice choosing the right style for you!

In the exercises, you'll practice selecting some of these styles for your own visualizations.

Create Your Free Account

By continuing, you accept our Terms of Use, our Privacy Policy and that your data is stored in the USA.