Creating Time Series Plots with the countries as subplots

Question

I am stuck when it comes to creating the plots as I don't know how to define the x and y variable. I was planning on making the countries as the x variable and then making the dates for the COVID cases as the y variable. I want to know what is the best way to do this. Furthermore, is there any way to plot each country in a different subplot?

Thank you for the help.

I have attached the following piece of code:

import pandas as pd

cases_raw = pd.read_csv(
filepath_or_buffer='https://raw.githubusercontent.com/aps1070-2019/datasets/master/APS_COVID_Jan22.csv',
index_col=0,
thousands=','
)



from matplotlib import pyplot
display(cases_raw)
df = pd.DataFrame(data=cases_raw) 
df.index.tolist() #returns the list of countries 
cases_raw.loc['Canada'] #733 cases

Screenshot of Problem

user_3281097 · Answer

First, let's modify the data to include columns with informative names and also let's use a better tabular representation by moving the dates to rows using .melt:

# Load data and name index as "Country"
df = pd.DataFrame(data=cases_raw).rename_axis("Country").reset_index()
# Move date columns as row values and keep "Count" as value
df = df.melt(id_vars=["Country"], value_vars=df.columns[1:], var_name="Date", value_name="Count")

Our df dataframe looks like this:

      Country        Date   Count
0     Afghanistan  2020-01-22       0
1         Albania  2020-01-22       0
2         Algeria  2020-01-22       0
3     Afghanistan  2020-01-23       0
4         Albania  2020-01-23       0
...           ...         ...     ...
2194      Albania  2022-01-22  246412
2195      Algeria  2022-01-22  234536
2196  Afghanistan  2022-01-23  159649
2197      Albania  2022-01-23  248070
2198      Algeria  2022-01-23  236670

Important: I just used the data from 3 countries to test quickly.

Now, we can group by "Country" and plot each group/country in a subplot by iterating on the axis of a subplot and the groups at the same time (using zip):

# Use nrows and ncols as you prefer, I know that I have 3 countries so...
fig, axes = plt.subplots(nrows=1, ncols=3, figsize=(12,4), sharey=True)
# Let's create groups by "Country"
country_groups = df.groupby("Country")
# Iterate on group keys (country names) and our subplot axis
for (country_name, ax) in zip(country_groups.groups.keys(), axes.flatten()):
    # Plot each group in the plot and include it's title
    country_groups.get_group(country_name).plot(ax=ax, title=country_name)
# Display plot
plt.show()

Result:

Creating Time Series Plots with the countries as subplots

1 Answers