Shekhar's Science Blog

Friday, April 30, 2021

Prediction-2021 for Confirmed Cases in India

 

Prediction-2021 for Confirmed Cases in India

 

Dr Himanshu Shekhar

 

 

Introduction

The second wave of CORONA is more dangerous and fatal. The COVID-19 with renewed energy and mutation has been playing havoc in India. The tremendous rise in number of confirmed cases and number of deaths have broken the moral of general population. On ground, the scarcity of bed, oxygen, medicine and courage has been deterrent to all positive energy. The cases have been continuously rising and I have been asked by many friends to take a look at the situation and predict about the turnaround. I tried to do the same taking data from website about the total confirmed cases in India.

Presentation of Raw Data



Raw data of total confirmed cases of India is plotted and it is visible that the rate of rise in 2021 is much higher than rate of rise in 2020. If 15 April 2020 is considered as start point, keeping 1000 cases as base and peak is considered around 15 September 2020, the duration in which turnaround is observed is around 150 days. The fall is also observed in around 150 days, maintaining the true nature of the exponential normal distribution. The optimized normal distribution is plotted for minimization of error till 15 February 2021 and the obtained optimised curve for the confirmed cases in 2020 in India superimposed on the actual data.



 

Nature of curve is almost matching to the actual situation. With such prediction capability of normal distribution curve. It is clear that the next batch will also follow similar nature with same slope of the upward and downward limb of the normal distribution curve.

 

Prediction

Considering the spread of pandemic as normal distribution curve with symmetrical rise and fall, the data is extracted for 2021 for prediction. Normal distribution curve is fitted and the square of error is minimized using solver to get the optimized prediction for the current total confirmed cases. The optimizer is using data up to 30.04.2021 and is using normal distribution curve with three parameters mean, standard deviation and peak number of cases. The nature of fit for the calculated duration is shown in the plot below.



To understand the real prediction, the data is generated for the entire curve and the situation is depicted in next plot. It is clear that peak is exceeding 7 lakhs and turnaround for the present situation occur on 29 May 2021.



 

 

Concluding remarks

If no accidental rise due to festival, marriages, religious functions, election or otherwise occurs, the current rate of rise predicts falling graphs to have less than 1 lakh daily confirmed cases by 23 July 2021. The correctness of data and deliberate reduction in number of daily testing may affect the prediction. This is a purely mathematical treatment of information. However, normal distribution curve represents the cases adequately. However, the case enhancer should not augment and bolster the current rate of rise of confirmed cases.

 

 

4 comments:

  1. Scary but we can prepare when we know and when we are aware. Excellent Analysis

    ReplyDelete
    Replies
    1. Thank you. Next post will be in depth analysis, using growth factor, doubling period etc.

      Delete
    2. Have u considered the vaccination impact which was missing in 2020?

      Delete
    3. No. Confirmed cases will not drop till herd immunity is achieved, say after vaccinating 60% of population, which is likely to occur after the second wave has gone. Vaccination can reduce severity after infection, only.

      Delete