How to Plot a Basic Pie Chart Of Taxa At A Given Taxonomic Rank Using Pandas And Matplotlib

Note

This post was migrated from my iNaturalist journal to my Jekyll blog on 2023-02-26. It was then migrated to my Quarto blog on 2024-07-27.

Suppose you have an exported CSV of iNaturalist observations, observations-<ID>.csv.

To follow this tutorial you will have to have Python installed, and the Matplotlib and Pandas packages installed. Pandas is not necessary, but it makes things convenient enough that I recommend using it here. In other contexts you may wish to plot pie charts without Pandas.

Assuming you have PIP installed, you can install Pandas and Matplotlib as follows:

pip install matplotlib pandas

Although, since Pandas actually uses Matplotlib as a dependency for plotting, it might suffice to simply use:

pip install pandas

Next, you must create a script file with the *.py extension. We can do fancier things with paths, but let us create the file pie_taxa.py using BASH.

touch pie_taxa.py

Now let us write some lines of code in pie_taxa.py. First we need to import the required packages.

import matplotlib.pyplot as plt
import pandas as pd

Next we can load our data using the pd.read_csv function, which assumes a CSV format by default. It has many other parameters, including changing the delimiter (see the docs), but we are fine with the defaults here.

import matplotlib.pyplot as plt
import pandas as pd
df = pd.read_csv("observations-<ID>.csv")

Now, as if by magic (but actually the hard work of software developers), we can create the plot in a single line of code. Let us do it for the kingdom level, which will require us knowing that this is represented by the taxon_kingdom_name column in our data file.

import matplotlib.pyplot as plt
import pandas as pd
df = pd.read_csv("observations-<ID>.csv")
df['taxon_kingdom_name'].value_counts().plot.pie()

There are a few things going on in the previous line of code. First is that df['taxon_kingdom_name'] has selected only the taxon_kingdom_name column. This is next passed to the value_counts() which counts the occurrences of each kingdom in that column and returns a Pandas series object with this information, and then we finally call the plot.pie method on this series object which… well… makes the pie chart.

If you run the code at this point you may be surprised to not actually see a plot appear anywhere. If you ran the code from the command line you might have seen something like <AxesSubplot:ylabel='taxon_kingdom_name'>. This is because creating the instructions of what goes on the drawing canvas is different from graphically rendering it. In order to do that, we can call plt.show().

import matplotlib.pyplot as plt
import pandas as pd
df = pd.read_csv("observations-<ID>.csv")
df['taxon_kingdom_name'].value_counts().plot.pie()
plt.show()

Running the above, you should see a window pop up. It has various settings for resizing, reshaping, zooming, and saving your figure.

What if you didn’t want to look at Kingoms, but rather orders, or families, etc? You would simply use a different column instead of taxon_kingdom_name. Here is a table of these similar columns:

Taxonomic Rank
`taxon_kingdom_name`
`taxon_phylum_name`
`taxon_subphylum_name`
`taxon_superclass_name`
`taxon_class_name`
`taxon_subclass_name`
`taxon_superorder_name`
`taxon_order_name`
`taxon_suborder_name`
`taxon_superfamily_name`
`taxon_family_name`
`taxon_subfamily_name`
`taxon_supertribe_name`
`taxon_tribe_name`
`taxon_subtribe_name`
`taxon_genus_name`
`taxon_genushybrid_name`
`taxon_species_name`
`taxon_hybrid_name`
`taxon_subspecies_name`
`taxon_variety_name`
`taxon_form_name`

Happy plotting.