#COVID19 projections will be hugely influential in the coming months as we prepare to slowly re-open society. Introducing @CovidProjection, a tool for epidemiologists/policy makers to track differences between models and for the same model over time
http://www.covid-projections.com


@CovidProjection consists of two components:
A data parser that downloads and compiles current and historical projection data.
An interactive dashboard that visualizes changes in projections over time


The data parser currently ingests data from two models:
@IHME_UW
@LosAlamosNatLab
Modeling pandemics is tremendously difficult - it’s far easier to be wrong than right. I applaud both groups for their efforts to make their data easily accessible in the public interest.


Modeling pandemics is tremendously difficult - it’s far easier to be wrong than right. I applaud both groups for their efforts to make their data easily accessible in the public interest.
The dashboard currently supports a historical view into @IHME_UW projections with more features to come.
In making this tool, I’m acutely aware that these graphs may be used inappropriately to promote immediate reduction in physical distancing/NPIs.
In making this tool, I’m acutely aware that these graphs may be used inappropriately to promote immediate reduction in physical distancing/NPIs.

Plots include a watermark because I want @CovidProjection to be tagged if these charts are used inappropriately. I plan to actively develop new features but may not release them publicly if the harm starts to outweigh the benefit. Consider yourselves warned. 


@CovidProjection was built in python
using @plotlygraphs Dash framework. In the interests of promoting active development, I’ve open-sourced the code here. https://github.com/yuorme/covid-projections


The original models can be found here:
@IHME_UW
http://www.healthdata.org/covid/data-downloads
@LosAlamosNatLab
https://covid-19.bsvgateway.org/#link%20to%20forecasting%20site

http://www.healthdata.org/covid/data-downloads

https://covid-19.bsvgateway.org/#link%20to%20forecasting%20site
This tweet is really starting to take off. For the record, I am not an epidemiologist and am actively trying to leave interpretation to the experts. Some of these experts you should follow are: @CT_Bergstrom, @reichlab, @nataliexdean, @mlipsitch, @BillHanage, @zbinney_NFLinj
And no, I don't have a soundcloud. If you're fortunate enough to have a stable income in these difficult times, please consider checking in on your neighbours, volunteering in your community, or making a charitable donation #cavabienaller
