Congratulations on a your pre-submission. The topic is of course relevant to the situation we are all going through and there is lots of data to work with. Your model selection is good and somewhat different to other groups (this is good). I am sure you will learn a lot trying to implement these models and exploring possible strategies going forward.
General comments:
Remove figure from the template notebook. It does not belong here.
Use inline citations for your reference list
Design + add rich picture as indicated in Section 2
WRT your three models:
ABM
There a lots of ready-to-use ABM models of COVID-19 that you can reuse. There is no need to implement an ABM from scratch, as long as you extend the model in some way. This could be by changing assumptions or exploring new simulated scenarios. I did a quick search and found these repositories:
Schelling
I can see you are having trouble running the Schelling model from mesa. I'll have a look and try to find the source of the problem. In the meanwhile there are other implementations that you can try to reuse/recycle. Here are a few:
Riot Model
You need to decide whether you will run this model using a "soup" of agents (meaning each agent can observe all the others), or if you will run it on a network (e.g., preferential attachment). In the first case the implementation is trivial, but not realistic. In the second case the implementation is a bit harder but much closer to reality. Depends on how much effort you want to put into it. Up to you. Have a look at this library:
Hi All,
Congratulations on a your pre-submission. The topic is of course relevant to the situation we are all going through and there is lots of data to work with. Your model selection is good and somewhat different to other groups (this is good). I am sure you will learn a lot trying to implement these models and exploring possible strategies going forward.
General comments:
WRT your three models:
ABM There a lots of ready-to-use ABM models of COVID-19 that you can reuse. There is no need to implement an ABM from scratch, as long as you extend the model in some way. This could be by changing assumptions or exploring new simulated scenarios. I did a quick search and found these repositories:
Schelling I can see you are having trouble running the Schelling model from mesa. I'll have a look and try to find the source of the problem. In the meanwhile there are other implementations that you can try to reuse/recycle. Here are a few:
Riot Model You need to decide whether you will run this model using a "soup" of agents (meaning each agent can observe all the others), or if you will run it on a network (e.g., preferential attachment). In the first case the implementation is trivial, but not realistic. In the second case the implementation is a bit harder but much closer to reality. Depends on how much effort you want to put into it. Up to you. Have a look at this library:
@JordanBoulous