search

ikashnitsky / covid19-nuts3

Replication materials for: COVID-19 in unequally ageing European regions
https://osf.io/abx7s/
MIT License
3 stars 3 forks source link

Why does Poland look so different? #1

Open mpascariu opened 3 years ago

mpascariu commented 3 years ago

Great graphic Ilya! And the code is just as nice 🥇

I was just wondering why the contrast between Germany and Poland is so strong since they are neighboring countries? Is it the data or the age structure of the populations? I understand that this pictures tells more about the ages of the populations than the C19 fatality.

Thanks a lot! Each time I look at one of you plot I tell to myself that I have to step up my game.

mpascariu commented 3 years ago

One way to understand more the covid-19 fatality using this map is to look at 1 age group at the time. Looking at it like this one can make more sense of the assumptions used which are pretty strong. Let me know your opinion.

map20

map40

map70

ikashnitsky commented 3 years ago

Hi Marius! Thanks a lot! This looks really interesting. Though, I think the focus on certain age groups changes the aim of the investigation a lot – using the same approach as we chose for the whole population you basically focus in the differences in the age distribution within the chosen 10-year age groups (I actually wonder how you produced these maps when the IFRs we used are only applied to 10-year age groups 🤔 ).

ikashnitsky commented 3 years ago

Did you ungroup the IFR estimates to single-years?

mpascariu commented 3 years ago

Hi Marius! Thanks a lot! This looks really interesting. Though, I think the focus on certain age groups changes the aim of the investigation a lot – using the same approach as we chose for the whole population you basically focus in the differences in the age distribution within the chosen 10-year age groups (I actually wonder how you produced these maps when the IFRs we used are only applied to 10-year age groups 🤔 ).

Hi Ilya, I am not looking at individual ages, but at 1 age group at the time (treated as a singe age). I was able to get the maps above only by filtering the data used in the map construction. Apart from that your and Jose's method stays the same.

I see the map you produced as an average across all age groups, I just separated the elements to be inspected one by one. Otherwise little can be learn of epidemic dynamic across Europe.

Did you ungroup the IFR estimates to single-years?

See the modified code here: https://github.com/mpascariu/covid19-nuts3

Thanks a lot!