Maharastra on 21.05.2020: Cumulative Confirmed Cases

Maharastra on 21.05.2020: Cumulative Confirmed Cases

Dr Himanshu Shekhar

Introduction: There have been continuous monitoring of daily and cumulative confirmed cases for Maharastra in various posts of the blogs. One prediction was made on 30.04.2020, which is reproduced below for further analysis.

Current Status: Although the prediction made on 30.04.2020 for India failed miserably, as an overestimated numbers and a fresh prediction is made on 20.05.2020, the same is not the case with Maharastra. The cumulative number of confirmed cases has been taken from Internet and Aarogya Setu and plotted against prediction.

It is clear from the above superimposed plot that although trend is similar, but numerical values of actual confirmed cases, as on 21.05.2020 is lower than predicted values of confirmed cases on 30.04.2020. Although, estimate is off, as seen from the superimposed curve, but indication of achieving 40000 cases on 21.05.2020 is marginally overestimated. The actual number of confirmed cases on 21.05.2020 is 39297. However, the Goodness of Fit using Pearson coefficient is ascertained for Maharastra from 30.04.2020 to 20.05.2020 and the value is obtained as 0.998355. This seems to be a good fit.

Analysis: It is irony of fact that Maharastra is still in line with the prediction made on 30.04.2020, while India has rejected the prediction, out rightly, by registering significantly less number of confirmed cases by 20.05.2020, needing a fresh prediction strategy. It is presumed that the graph for actual is deviating downwards and it is expected that by the end of May 2020, Maharastra will also register significantly less number of cases than prediction made on 30.04.2020, needing a fresh look at the prediction strategy and mathematical treatment.

The variation of prediction against actual cases is superimposed over equi-inclined linear variation. It is clear that the data is slightly moving upward indicating more weightage to data on y-axis. This indirectly means that data on y-axis, or prediction by Shekhar on 30.04.2020 is on higher side and the values can have further deviation with time. A histogram is also generated with date wise value of cumulative confirmed cases. The numerical value of actual confirmed cases is given on plot and the predicted data against actuals is seen in the plot, with similar conclusion as drawn from the linear variation plot of the two variables. The prediction is always on higher side as compared to the predictions made on 30.04.2020.

Conclusion: The prediction versus actual number of cumulative confirmed cases for Maharastra is re-looked in light of the milestone of achieving 40000 cases on 21 May 2020. The milestone slipped by a small amount, but it gave hope that the actual cases will continue to be much lower than predictions in future for Maharastra and in next 10 days the prediction will be invalid as overestimated value for cumulative confirmed cases in Maharastra. However, the prediction for Maharastra made on 30.04.2020 is valid for 21 days, which is higher than the validity of prediction for India. Obviously, the nature of Maharastra is better understood by author. Overall, hope is that prediction will fail as overestimate and some control will prevail in Maharastra in near future. 

Dr Himanshu Shekhar