Mapping and visualizing the spreading of COVID-19

Valentina Alto

Follow

Mar 14, 2020 · 4 min read

On 23 January 2020, the central government of China imposed a lockdown in Wuhan and other cities in Hubei province in an effort to quarantine the epicenter of an outbreak of coronavirus disease 2019 (COVID-19).

However, given the current mobility and globalization, it was inevitable for this disease to expand also outside its epicenter, resulting in what the WHO defined as a real pandemic.

Many countries these days are experiencing restrictions comparable to those in Hubei. Yet this is the only weapon (at least so far) we have to fight the further spreading of this disease. And the exceptionality of the situation can be understood even more deeply if we look at the numbers of this crisis.

In this article, I’m going to provide a graphical representation of the evolution of the COVID-19, relying on this real-time updated dataset. For this purpose, I will use the python library Plotly, and I will provide global time-series data as well as country-specific information.

Let’s first download the dataset and do some data cleaning (specifically, I converted the column ‘Date’ to datetime).

import pandas as pd
df = pd.read_csv("covid_19_clean_complete.csv")
df['Date'] =  pd.to_datetime(df['Date'])
df.head()

Let’s now examine the spreading of the disease around its epicenter. So let’s first filter our dataset:

m = [df['Country/Region'][i]=='China' for i in range(len(df))]
df_china=df[m]
df_china.head()
df_china=df_china.groupby('Date').sum().reset_index(drop=False).sort_values('Date')

Now let’s plot the result on a map, using ‘Asia’ as scope.

import plotly.express as px
fig = px.scatter_geo(df_china, lat='Lat', lon='Long', color="Province/State",
                     hover_name="Province/State", size="Confirmed",
                     animation_frame="Date", size_max=30,
                     projection="natural earth", scope='asia')
fig.show()

As you can see, the biggest bubble (where the size of the bubble indicates the number of infected) is in Hubei, the province of the epicenter.

We can also inspect which was the pace of the spreading, taking the trace of infected, recovered and deaths.

import plotly.graph_objects as go
import pandas as pdfig = go.Figure()
fig.add_trace(go.Scatter(
                x=df_china['Date'],
                y=df_china['Confirmed'],
                name="Confirmed",
                line_color='deepskyblue',
                opacity=0.8))fig.add_trace(go.Scatter(
                x=df_china['Date'],
                y=df_china['Deaths'],
                name="Deaths",
                line_color='tomato',
                opacity=0.8))fig.add_trace(go.Scatter(
                x=df_china['Date'],
                y=df_china['Recovered'],
                name="Recovered",
                line_color='forestgreen',
                opacity=0.8))# Use date string to set xaxis range
fig.update_layout(title_text="COVID-19 spreading in China")
fig.show()

The explosion of this virus was exponential. Indeed, virologists attributed to this new virus a basic reproductive number of 2–3. It means that, assuming that everyone is susceptible to contracting the virus, each infectious person transmits the disease, on average, to 2–3 people.

Interestingly, by observing the pace of this spreading across different countries, I noticed a common exponential pattern, which eventually plateau (generally when lockdown policy is enforced by governments). In particular, I examined Italy, Iran and the US, obtaining the following results:

Now let’s see the overall evolution across countries:

import plotly.express as px
fig = px.scatter_geo(df, lat='Lat', lon='Long', color="Country/Region",
                     hover_name="Province/State", size="Confirmed",
                     animation_frame="Date", size_max=30,
                     projection="natural earth")
fig.show()

import plotly.graph_objects as go
import pandas as pddf = df.groupby('Date').sum().reset_index(drop=False).sort_values('Date')fig = go.Figure()
fig.add_trace(go.Scatter(
                x=df['Date'],
                y=df['Confirmed'],
                name="Confirmed",
                line_color='deepskyblue',
                opacity=0.8))fig.add_trace(go.Scatter(
                x=df['Date'],
                y=df['Deaths'],
                name="Deaths",
                line_color='tomato',
                opacity=0.8))fig.add_trace(go.Scatter(
                x=df['Date'],
                y=df['Recovered'],
                name="Recovered",
                line_color='forestgreen',
                opacity=0.8))# Use date string to set xaxis range
fig.update_layout(title_text="COVID-19 spreading worldwile")
fig.show()

Besides the difference in numbers, the initial behavior looks pretty similar. The main difference is the moment of plateauing: in China, since it started way before, the restrictive measures started producing effects from the middle of February, more or less when other countries started experiencing the first positive cases.

Hopefully, all the measures taken by countries and citizens should lead to the same result as China, yet only numbers and time will tell us.

References: