We can semi-reliably predict (tuning required) the trend of new cases due to COVID-19 in a given district using standard Time Series Analysis methods such as Holt-Winters or ARIMA.
forecast.R
contains the code for making a training and testing a model and finally applying it on available data
to forecast the near future.
But before that, we need the data, the scripts for doing that are in the Makefile
, but you don't need to know how
it works, generate the data for your district using:
make district='"New Delhi"'
Keep in mind, the data is from covid19india.org
and is updated daily, to download a new copy of the data, run
make fetch
make district='"New Delhi"'