search

ctedja / vaccinations_and_regressions

A work in-progress project analysing vaccination trends in the US @MoctarAboubacar
MIT License
0 stars 0 forks source link

Loading Shapefile #3

Open ctedja opened 3 years ago

ctedja commented 3 years ago

Reckon you can separately send me the shapefile you're working with? I suspect that you may be working with more than just the .shp file but that's all that was loaded/pushed onto GitHub. Then I'll add it manually into my directory :)

@MoctarAboubacar

MoctarAboubacar commented 3 years ago

Hey yes I can send this will do! In fact I am literally only using that single .shp file, but there are the other data that come with it, I'll upload those too!

2021년 5월 5일 (수) 22:40, Clinton Tedja @.***>님이 작성:

Reckon you can separately send me the shapefile you're working with? I suspect that you may be working with more than just the .shp file but that's all that was loaded/pushed onto GitHub. Then I'll add it manually into my directory :)

@MoctarAboubacar https://github.com/MoctarAboubacar

— You are receiving this because you were mentioned. Reply to this email directly, view it on GitHub https://github.com/ctedja/vaccinations_and_regressions/issues/3, or unsubscribe https://github.com/notifications/unsubscribe-auth/AM4DL5RJDEX66JEJRGPH7KTTMFRGRANCNFSM44FHIUVQ .

ctedja commented 3 years ago

Okay thanks! Looking forward to it.

As for the vaccinations dataset, I think I downloaded my first on 17 April -- you too? If so, I can download another again on 17 May and we can get start to have monthly snapshots to gradually begin comparing rates :)

ctedja commented 3 years ago

https://github.com/nytimes/covid-19-data/tree/master/mask-use Oh!

MoctarAboubacar commented 3 years ago

Hey sorry I've been AWOL on this recently. That's great news, looks like a useful dataset!! Let's schedule a call to discuss the whole project? Say Friday?

On the regression side I gave it a little more thought and think we need to modify our dependent variable. The issue is that we are trying to model a bounded variable (a percentage) which is not supposed to be done with vanilla regression : normally we would do a tobit regression or beta regression...both of which are pretty complex and I have not done before.

Solution is actually realy really simple and more elegant than what we wanted to do: we use the raw number of people vaccinated as the dependent variable and then include the state population as an independent variable. So controlling for size of population we can see "how many extra people" are vaccinated for the change in level of our other independent variables.

On another front I was only able to find state level data for the evolution of vaccination. Let's talk about it, but it probably won't help our county level regressions much. The insights are cool though!

2021년 5월 18일 (화) 19:17, Clinton Tedja @.***>님이 작성:

https://github.com/nytimes/covid-19-data/tree/master/mask-use Oh!

— You are receiving this because you were mentioned. Reply to this email directly, view it on GitHub https://github.com/ctedja/vaccinations_and_regressions/issues/3#issuecomment-843119410, or unsubscribe https://github.com/notifications/unsubscribe-auth/AM4DL5Q7HAEPXJYTTHTANNDTOJLHBANCNFSM44FHIUVQ .