Simulating the effectiveness of lockdown using CLUBASID

Olusola I Olaoye
6 min readOct 4, 2020

If you have been living on earth this year, you must have heard of lockdowns going on in different parts of the world, due to the Covid-19 pandemic, in order to reduce the spread of the virus. One of the main question people have been asking concerning the lockdown is how effective they are. I mean, we can all agree that it is very inconvenient for many people, but can we simulate how effective they really are? How big of a difference does an effective lockdown play in terms of curbing the spread of the virus. I will be using CLUBASID software (Cluster Based Simulation of Infectious Diseases), which is a tool that helps simulate the spread of infectious diseases. I want to simulate for 3 different population clusters. In the first cluster, people are simply moving around randomly. In the second cluster, people are heading towards different locations around that same cluster. In the third cluster, people only go to the closest location to their home (the lock-down rule). In our simulation, we are representing 1 second in real-time as 1 day in simulation time.

Screenshot of CLUBASID

The above image is the view of the software when you first open it, so let us begin. We start by creating a new simulation.

CLUBASID app

Looking at the image below, you can see that I ticked only the “droplets” checkbox, under the “Mode of transmission” header. This means that the disease spreads through droplets alone (You can of course add other modes of transmissions depending on the disease you are simulating for). You could also choose to adjust latency period, incubation period, etc. Please note that covid-19 is reported to spread through direct and indirect contacts as well.

After clicking the “use virus settings” button shown above, a “cluster container” shows up like in the image below. This shows the bounds that individuals in this particular cluster are restricted to.

To make sure you are satisfied with the virus characteristics, click on the disease info button on the top right. This is a window that pops up to show you the characteristics of the disease we are trying to model.

Now that we have initialised the virus settings, let us set some more values. First, tick the multi-cluster simulation box (at the top of the left panel). These are some of the values I changed (on the left panel).

  1. Tick the multi-cluster simulation checkbox.
  2. Increase the world width and world height to approximately 500.
  3. Increase the initial infection rate slightly above 1%.
  4. leave restriction type as completely free.

After you do the following, click on “add population” button below. I named this cluster “free movement”.

Now drag the second cluster so that it does not overlap with the first cluster, otherwise, you will get a warning like the one below.

Leave all the values the same as the first cluster, but this time, change the restriction type to public locations. I set up the number of public places to 10. Then add this cluster and set the name to “public places”.

For the last cluster, change the restriction type to “Closest location/ lockdown” option. leave the number of key locations at 10. Now add cluster, enter the name “Lockdown”, then press simulate.

Just after 3 days (in simulation time), you can see “sub clusters” stating to form in the lockdown cluster because people are only going to the closest location to them. So what we get in this case is some sub-clusters will be infected and some will not be. If you are in a sub-cluster that has no infected person, your chances of getting infected is non-existent. However if you are in a sub-cluster that has an infected person, you are constantly exposed to such individual, hence your chances of contacting the virus is high.

After 15 days of simulation time, we can see that things are not looking good for the cluster on the top right. This is because people are going to different locations all over that cluster. This means an infected person has access to everyone in that cluster (given the right time) because he/she can simply visit any location in that cluster

After 29 days of simulation time, people are stating to recover (you can see the green dots showing people who have developed immunity for the virus). Notice something different in the cluster at the button (which is the lockdown cluster). You can see that there are not too many green dots to show people recovering from the virus.

After 76 days in simulation time, everyone who had the disease has recovered and we can see the epidemic curve (showing infections) on the graph. Notice how the green dots in the lockdown cluster below isn’t as much (compared to the other 2 at the top). Showing that lots of people didn’t have the disease to begin with.

So after 76 days we have had 53 deaths, and 1278 recoveries. However, what this shows us is the global data which means the data for the 3 different clusters combined. We want to know how each cluster performed in terms of infection spread (given that they had different restriction types). The question is, what role did the restriction types play in the spread of infection?

So we change from global space to local (at the bottom right). We then select the “free movement” cluster and update the graph. The image below shows the area under the graph.

We do the same for the next cluster “Pub locations” and the image below shows what the area under the graph looks like.

Finally we do the same for the third cluster called “lockdown” and the image below shows what the area under the graph looks like.

In conclusion, we can see that the area under the graph for the final cluster where lockdown is implemented is the smallest, demonstrating an effective lockdown. Note that the area under the graph does not mean number of unique infections. Number of unique infections will be calculated by adding up the number of current infections, the number of recoveries and the number of deaths.

This application is available for free on the Microsoft Store : https://www.microsoft.com/en-us/p/pandemic-simulator/9n86x1xhz765?activetab=pivot:overviewtab

--

--