To follow on from my #India #Coronavirus #epidemic tracker I post some charts in a follow-on thread that paint a data-picture of the epidemic in the States 
The epidemic is still growing in all the states.
The RQ - Recency Quotient - i.e the proportion of cases that occurred in the last 14 days varies across the states but many are at very high levels. 3 states exp - Karnataka, Andhra and Bihar are above 50%. Delhi is the lowest.
fatality rate is a real problem. How do we define it? and what's the best measure given we dont really have a case-registry database. The Govt uses the crude ratio of deaths to cases. But it is flawed, given that 38% of cases are recent (last 14 days) infections.
One could use a revised definition: Deaths div by so-called closed cases (i.e Cured + deaths). this measure gives the following chart:
Finally, my preferred definition is to use a modified denominator - the cases that had occurred till a date 14 days ago. This gives the following chart:
But all three measures are bedevilled by the known problem of under-reporting of deaths. But then so are cases going undiagnosed and unreported. Messy problem! Lastly, this is a scatter plot of States - pon incidence by a revised measure of case fatality. Interesting spread!
