search

CoronaWhy / task-geo

"Help us understand how geography affects virality."
MIT License
10 stars 20 forks source link

Variational SEIR modell #22

Open sammerpuranpost opened 4 years ago

sammerpuranpost commented 4 years ago

Creating a SEIR modell to fit it to cases per capita vs time or deaths per capita versus time. Datasets needed:

Timeseries of cases per country or deaths per country

Technical approach: Create a SEIR modell, which with defined params and fit it with data using least squares.

To do: Fit the modell using real data to update the params. Then predict the evolution of cases using the modell

hyberson commented 4 years ago

Please include the URL https://trello.com/c/fdGGQzxh in the issue description. Discuss organizational aspects only in the Trello card –https://trello.com/c/fdGGQzxh.

sammerpuranpost commented 4 years ago

The model is here https://colab.research.google.com/drive/1mWB38jVZ2yjj5ENyRnOMU2yiLXSrMm3w#scrollTo=5zZ0Uvnw6686 However currently the data are csv files and I would like to use the data like it is intented in the repo, can someone help me adapt the code so that it can use data in our repo? The data I am currently using is https://data.humdata.org/dataset/novel-coronavirus-2019-ncov-cases