Corosim – historical estimates and model predictions for Covid-19 in Finland | Hacker News Reader

Corosim – historical estimates and model predictions for Covid-19 in Finland

corosim.fi

25 pointsbaobabKoodaa8 comments

This is a pro bono project that we built on top of Epidemic Calculator. We thought Epidemic Calculator was awesome as a visualization and modelling tool, but it was missing real data and the ability to do "real" scenarios of an ongoing epidemic. So we brought historical estimates into it (and a bunch of other stuff).

I think you should put "this is just a toy" in bold text on top.

The real work is in doing a good job of fitting the model parameters, but you leave this, the most crucial work, as comments in the source on some other page? : /

What? This is not "just a toy". And we did not leave out the work of choosing parameter values - we made a large effort to select reasonable default values for all the parameters. Also, what are you referring to with this "comments in the source on some other page"? Which page? Which comments?

> [...] our historical estimates are simple "back of the napkin" type calculations. We refer to the source code for details on historical estimates.

The exact parameter values together with their uncertainties makes a serious model.

Providing a set of differential equations with 11 mysteriously chosen values makes it little more than a toy.

> > [...] our historical estimates are simple "back of the napkin" type calculations. We refer to the source code for details on historical estimates.

> Providing a set of differential equations with 11 mysteriously chosen values makes it little more than a toy.

I'm guessing you think that historical estimates were created with the model, by running the model with super secret 11 parameters? This is not the case. Historical estimates were not created with the model. Historical estimates were created with simple "back of the napkin" type calculations, like it says in the text you quoted, and the calculations are linked right next to that text.

Also, what exactly is mysterious about the chosen parameter values? For example, if you click the question mark tooltip near R0 parameter, you will get a detailed description about how we calculate the default value for it. What is mysterious about it?

Original Epidemic Calculator is here: https://gabgoh.github.io/COVID/index.html

Can you say more about your development process? the model building and the GUI building.. components, your decisions.. etc

The model in Corosim is almost entirely the same as Goh's model in Epidemic Calculator. We made a lot of changes to the GUI, mostly motivated by a desire to answer practical questions related to the epidemic we have (as opposed to theoretical questions related to a theoretical epidemic). Have you checked the text under "Differences between Epidemic Calculator and Corosim"?

I'd be happy to answer any questions you have if you want to narrow down the scope a little bit. Like, maybe you are wondering why x axis zoom was changed, or maybe you are wondering why we removed "severe at home" state from the model? Or maybe you looked at the code and you want to discuss the extensive refactoring that we did?

Reply on news.ycombinator.com