Manindra Agrawal Profile picture
Apr 8, 2021 159 tweets 43 min read Read on X
@stellensatz @Sandeep_1966 @Ashutos61 @shekhar_mande Starting a separate thread on district level predictions. The predicted trajectories do not match as well as for states because of smaller population.
Let us start with Pune -- it was earlier posted on thread on states. It remains on track to peak during April 12-15 at around 11.5K infections/day. Image
Next Mumbai where infections are increasing quite fast. The rise will continue for another couple of weeks and likely to peak during Apr 21-25 at around 13K infections/day. Image
Bengaluru will peak even later: during Apr 26-30 at around 8K infections/day. Image
Finally Lucknow. It should peak during Apr 20-25 at ~3.3K infections/day. I am tracking a few more cities of UP like Kanpur and Prayagraj. Rise there is very recent and not yet captured by the model. Image
<Update on 10/4> Kanpur and Prayagraj are now captured by the model. But let us start with updates. First Pune. Earlier prediction stays. The blue curve has started flattening, an indication of peak. Image
Next Mumbai. Peak gone up slightly to 14K. To avoid any confusion, I have now explicitly noted in plot that new infections are 7-day average. Image
Bengaluru's blue curve continues staying on the orange curve without any change. Image
Lucknow peak has gone up slightly to ~3.5K infections/day and its timing remains the same. The blue curve appears closing in on the orange one. Image
Prayagraj started rising very fast but has slowed down a bit now. It is expected to peak at around 3K infections/day during April 16-20. It is early days though and peak value is likely to change. Image
And finally Kanpur. Blue curve right on orange one! Clearly Corona is a considerate virus, making sure that it follows the trajectory in the city of SUTRA's origin 😊. Peaking at ~900 infections/day during April 16-20. Image
Forgot to include Varanasi. It is expected to peak at ~900 infections/day during next five days. The blue curve took a Holi break here as well😀. Image
<Update on 14/4> Pune has peaked and now hovering. Hopefully, it will start to come down in a few days unlike Punjab!😊 Image
Mumbai is close to the peak. Both this and Pune curve support the prediction that Maharashtra is now peaking. Image
Bengaluru curves are moving in sync. Peaking during Apr 25-30. Image
Adding Chennai. It is at start of the rise, and blue curve may take a different direction. Let us see where it goes. As of now, peaking in May. Image
Now to cities of UP. Lucknow is seeing maximum number of cases. Blue curve has higher slope, but is it tapering off? Current peak value of ~5K cases during 15-20 will depend on this. Image
Prayagraj saw a very sharp rise but is now tapering off. Orange curve is still predicting a sharp rise with peak to come soon. Image
In Kanpur the two curves continue to move in sync. Peaking during Apr 20-25 around 1.5K. Image
Finally Varanasi. It also has both curves in sync. Peaking during April 15-20 at ~2K. Image
Adding Raipur, a city that is seeing very sharp rise. Orange curve has faithfully following blue one, so hope that blue will reciprocate in future 😀. Peak is round the corner, within a week. Image
<Update on 16/4> Pune is past the peak and is now on way down. I am zooming out to show the entire timeline. @stellensatz Image
Mumbai continues to flatten and is following orange curve closely. Image
Bengaluru's blue curve has become jittery. Making up its mind whether to follow orange one?🤔If it decides to follow orange curve, then peak will arrive during April 25-30 at ~11K. Image
Lucknow curves are in sync now. Peak scheduled during April 20-25. Image
In Kanpur both curve continue in sync. Peaking during April 20-25. Image
Varanasi curves also in sync. Peaking in next few days! Image
Raipur is also very close to peak. Coming in next few days. Image
Adding Korba, another city in Jharkhand that saw sharp rise. Peaking during April 20-25. Image
Also adding Noida. Seems to be peaking during April 20-25. Blue curve is a bit erratic so estimates may change. Image
<Update on 18/4> @stellensatz Pune is a bit up-and-down. We will just have to wait and see where it goes. Hopefully, parameter values are not changing! Image
Mumbai has almost flattened. Image
Two days ago, Bengaluru curve was wondering whether to stay on orange or not. Looks like it decided against it.😐Revised trajectory will need some more data. Image
Adding Thane. It seems to have peaked before predicted time. Let us see how it goes. Image
Adding Patna. Both curves are in sync. Peaking in next few days. Image
Nagpur was drifting earlier, and seems to be stabilizing now. I would not place too much faith on predictions yet. Image
Raipur has hit the peak. Image
Korba is nearing peak. I mistakenly wrote last time that Korba is a town in Jharkhand -- it is in Chhattisgarh. Image
Adding capital of Jharkhand -- Ranchi. It appears very close to peak. Image
Chennai is moving up gradually compared to many other cities. Is blue merging with orange? Image
Now to cities of UP. First Lucknow. Is it peaking before projected? Image
The two curves continue to be in sync in Kanpur. Peaking within a week. Image
Curves for Noida are also in sync now. Also peaking within a week. Image
Has Prayagraj hit the peak? Image
Finally, Varanasi curve has bent, so seems to be at peak. Image
<Update on 20/4> @stellensatz Pune continues to hover. Making up its mind where to go?🤔 Image
Mumbai curve is bending before predicted. We do not mind it😊 Image
Thane continues its downward journey... Image
Bengaluru curve had decided to go another way two days ago, but seems to have changed it mind and is coming back to orange one😊 Image
Chennai curves seem in sync now - will wait for a few more days to confirm. Image
Adding Kolkata. The two curves are in sync. Peaking during May 5-10. Image
Patna has overshot the peak. Let us wait for a few days to see how is it going. Image
Ranchi has come very close to peak... Image
Raipur is clearly turning... Image
Korba seems to be turning a little earlier. Let us see how it goes in the next few days. Image
Lucknow curve is going in a different direction. Will it continue to bend?🤔 Image
Varanasi has also overshot the peak. Will it turn? Next few days will tell. Image
Kanpur curves still moving in sync with blue peeing out just a little. Image
Noida curves are in sync. Peak expected during April 26-30. Image
Prayagraj is at the peak and hovering a bit. Image
I am tracking some more cities: Ahmedabad, Surat, Bhopal, Indore, and Nagpur. However, their trajectories have not stabilized and a few more days are required to make prediction with some confidence.
<Update on 22/4> @stellensatz Starting with Pune as usual. Blue curve continues to hover. Note that the plot is for Pune district -- I understand Pune city is reducing. Image
Mumbai continues downward... Image
... and so does Thane. Image
Adding Nagpur that has now converged. It should peak within a week. It has gone through an unusual trajectory as can be seen below. Some other districts also show similar trajectory. Image
Bengaluru curves continue to play hide-and-seek. It hard to develop much confidence in projections until they move is sync for a few days continuously. Image
Kolkata curves are in sync. Peak shifted to May 10-15. Image
Chennai curves are close but have different slopes. Let us see when they converge. Image
Adding Ahmedabad that seems to be converging now. If they do, peak to arrive around 30th April. Image
Patna presents a special case. The two corves were moving in sync but blue has continued to go up while orange turned. This could due to either significant number of cases from outside or start of a new phase. We will know which soon. Image
Adding Bhopal. It also has a shape similar to Nagpur. Image
Another similar case is Indore. Both Bhopal and Indore are a little jittery and we need to observe for some days to see if the curves move in sync. Image
Now to some districts in UP. Lucknow first. The blue curve appears to be at peak. Hope it starts going down soon! Image
Prayagraj curve is also hanging around the peak. Will it turn downwards soon? Let us hope so! Image
Kanpur is almost at the peak now. Image
Varanasi, like Patna, has overshot the peak. Possible reasons are the same. Image
Noida tracking well. Peaking during April 25-30. Image
Moving east, Korba appears to have hit the peak. Image
Raipur is past the peak and on the downward journey. Image
Finally Ranchi. Blue curve is peeking out at the top but has turned. However, it turned like this a few days ago also. Will it continue to turn?🤔We will know in a few days. Image
I am working on plots for top 50 districts (large number of active infections or large increase in active infections) of India and plan to put them on a website soon. Watch this space.
<Update on 24/4> @stellensatz Pune still hovering... Image
Mumbai still going down... Image
Thane still going down... Image
Nagpur nearing peak... Image
Bengaluru still uncertain... Image
Kolkata on track... Image
First change is Chennai. The two curves appear to have converged. Peak during May 15-20. Image
Patna continues its upward journey! I understand there is a large influx of people from outside there. We have not modeled that, so predictions are off. Still, it is expected that in a few days the blue curve should turn downward. Image
Bhopal curve has flattened suddenly. Possibly a sign of reduced testing? Image
Same is the case with Indore. Image
Lucknow has turned downward. Hope it continues... Image
Prayagraj continues to confound. It seems on a different trajectory now. Is it a phase change? We need a few more days data to be certain. Image
Kanpur is hitting the peak now. Will it turn, or will go on a different trajectory like Prayagraj? Image
Varanasi has finally turned! Image
Noida curve, which had flattened yesterday, has resumed upward journey. Such intermittent flattening is likely due to delay in testing results. Image
Korba is now hovering at the peak. Image
Raipur is continuing its downward journey. Image
And finally Ranchi. It had shot up above the orange curve peak but over past two days has turned and seems to be at the peak now. Image
<Update on 26/4> @stellensatz Pune has started going down again. Is there another twist in the future? Image
Mumbai is well and truly on downward journey. Image
Thane curve is flattening a bit, however, it is likely to be temporary. Image
Nagpur is going off a bit near the peak. Image
Bengaluru continues of confuse. The parameters continue to drift and so it is hard to estimate their correct values. Image
For Kolkata, on the other hand, parameter values are very stable. Image
Chennai continues on the predicted trajectory. Image
Patna has turned! Is it for good? We will know soon. Image
Bhopal flattening continues. I am told it could also be because of lockdown imposed sometime ago. Image
Indore moving identically. Image
Lucknow is now hovering. Image
Prayagraj curve has turned downward! So the trajectory change around the peak may not be due to a new phase, but because of delayed testing results. Image
Kanpur is off on the similar trajectory as Prayagraj. Is it also due to delated testing results? We will know in a couple of days. Image
Varanasai has started hovering. Image
Noida trajectory is jerky. Testing delays? Image
Korba continues to hover. Image
Raipur is well and truly on downward trajectory. And it is coming down as rapidly as it went up. Image
Ranchi has turned up after flattening at the peak. Where does it want to go?🤔 Image
Adding Ahmedabad. I now have some confidence in predicted trajectory but there is still uncertainty. Orange curve peaking during May 1-5. Image
<Update on 29/4> @stellensatz Delayed by a day due to some pressing engagements... Pune is well-and-truly on its way down now! Image
So is Mumbai. Image
And Thane. Image
Nagpur is at the peak. Image
Bengaluru parameters have stabilized now. Peaking during May 5-10. Image
Kolkata continues to follow projected trajectory. Image
Chennai is diverging a bit. Image
Patna appears to have turned for good! Image
Bhopal plot has been updated with more data points as it was a little unstable. It seems close to peaking. Image
Same for Indore. The two MP districts are moving almost identically! Image
Lucknow on its way down now! Image
Prayagraj well on its way down! Image
Kanpur may be turning now... Image
Varanasi, after hovering for some time, appears on its way down. Image
Noida shooting up! Hopefully it will turn soon like some other districts. Image
Korba is still hovering. Image
Raipur is reducing fast. Down by more than 40% from peak value. Image
Ranchi is still going up. Hope it turns soon. Image
Ahmedabad curve updated. It is stable now. Peaking time remains same: May 1-5. Image
Adding Surat. It is close to peaking. Image
<Update on 2/5> @stellensatz I bring mostly good news today! Pune continues downward journey with zig here and a zag there. Image
Mumbai, a city with little time to waste, is going straight down. Image
Thane, initially influenced by Pune, is now learning from Mumbai. Image
Nagpur is just past the peak. Image
Bengaluru stays on track to peak during May 5-10. Image
Kolkata back with orange curve. Peaking with Bengaluru. Image
Chennai is on a relaxed journey with no hurry to peak or rise. Image
Ahmedabad may have peaked. Image
Bhopal also may have peaked. Image
And Indore will not want to be behind Bhopal I suppose. Image
UP districts are taking a leisurely way down after peaking. Lucknow: Image
Prayagraj: Image
Kanpur: Image
Varanasi: Image
The only exception is Noida. It is is no mood to peak despite repeated invitations from orange curve. In a hurry to go somewhere?🤔 Image
Moving east. Patna also seems to have peaked and meandering downward. Image
Raipur learning from Mumbai!😳 Image
And Korba learning from Chandigarh!😳Chhattisgarh seems a really cosmopolitan state. Image
Ranchi may be coming down now. Image
Finally, Surat has peaked as well! Image
I forgot to add Khorda from Odisha. The district containing city of Bhubaneswar. It appears near the peak. It appears restrictions were imposed whose effect is visible since past one week. This will delay the peak a bit. Should arrive in next few days. Image

• • •

Missing some Tweet in this thread? You can try to force a refresh
 

Keep Current with Manindra Agrawal

Manindra Agrawal Profile picture

Stay in touch and get notified when new unrolls are available from this author!

Read all threads

This Thread may be Removed Anytime!

PDF

Twitter may remove this content at anytime! Save it as PDF for later use!

Try unrolling a thread yourself!

how to unroll video
  1. Follow @ThreadReaderApp to mention us!

  2. From a Twitter thread mention us with a keyword "unroll"
@threadreaderapp unroll

Practice here first or read more on our help page!

More from @agrawalmanindra

Apr 23, 2023
@stellensatz I am back with Covid-19 predictions, when it already seems to be peaking! Reason for being so late is that the model was unable to capture the trajectory due the numbers being very small (5-6K per new for India is nothing).
Once the numbers crossed 10K per day, the model did capture the trajectory somewhat. By 15th April, this is what it predicted: Image
The predicted peak was around 50K new cases per day somewhere in mid-May. I made a couple of media comments also based on this. However, before I could post the predictions on twitter, the trajectory changed again.
Read 7 tweets
Dec 21, 2022
@stellensatz @GyanCMehta The rapid spread of Omicron in China over the past one month has raised several questions:
1) Why is it happening in China now after such a long time despite vaccination? [1/18]
2) Numbers in some countries are also rising. Is it likely to spread to other countries too?
3) Should we be concerned in India?

I recently did a simulation of spread in China and a few other countries using SUTRA. Here are the conclusions. [2/18]
1) The new infections plot for China since March this year when a major outbreak happened. [3/18]
Read 18 tweets
Jan 5, 2022
@stellensatz The assumption I made - India will behave similarly to SA - turns out to be wrong.
Indian trajectory is rising faster than projected earlier. We now have enough data to start doing preliminary predictions (as opposed to projections).
Maximum data is from Mumbai. Phase plot still showing a drift indicating that parameter values are likely to change. Image
Current parameter values suggest a peak around 15th Jan. Need to wait for a few more days for parameters to stabilize (of course, if lockdowns are imposed, it will change the parameters again). Image
Read 42 tweets
Dec 22, 2021
@stellensatz South Africa peaked on 17th December. Slightly before our prediction . With data up to 16th Dec, SUTRA prediction of trajectory matches well so far. Image
There seems to be broad agreement that Omicron arrived in October in SA. If true, our earlier assumption that Omicron arrived in August causing a jump in contact rate is incorrect. So how did Omicron change parameter values?
A new phase started in SA from November 1 and stabilized after three weeks. Value of contact rate (beta) increased from 1 to 1.35 and reach (rho) increased from 0.86 to 1.04.
Read 20 tweets
Dec 3, 2021
@stellensatz More information on omicron variant. First, a more careful SUTRA simulation shows that contact rate (beta) jumped to 1 by Aug-end. It strongly suggests presence of a new mutant that was active by Aug.
Note that the initially numbers would be very small and so genome sequencing may not throw up any case, especially if (as has been reported) most of the cases are mild and thus may go unreported. So it is not a surprise that first case was reported in November.
Why was the initial spread slow given that it is a highly infectious mutant?

Reach of the pandemic from July until now is around 85% (this is computed using the serosurvey that showed ~47% seropositivity in May '21). Calibrated model shows that natural immunity in Sep was ~77%.
Read 15 tweets
Oct 29, 2021
@MenonBioPhysics gave an good critique of SUTRA model a few days ago (see ). It makes many important points regarding epidemiological modeling. In this thread, I am going to develop his analysis further and show what is the purpose of SUTRA model.
A key point made is that epidemiological models stratify population on age basis since interactions vary widely with age. Further stratification is done between mild and seriously ill cases. Former is useful to make policy decisions about controlling the spread.
Latter is useful in predicting hospital beds requirements. These advantages, however, come with some complications. More compartments mean more interactions, which mean more parameters to estimate. Epidemiologists use mobility, social structures etc to estimate their values.
Read 25 tweets

Did Thread Reader help you today?

Support us! We are indie developers!


This site is made by just two indie developers on a laptop doing marketing, support and development! Read more about the story.

Become a Premium Member ($3/month or $30/year) and get exclusive features!

Become Premium

Don't want to be a Premium member but still want to support us?

Make a small donation by buying us coffee ($5) or help with server cost ($10)

Donate via Paypal

Or Donate anonymously using crypto!

Ethereum

0xfe58350B80634f60Fa6Dc149a72b4DFbc17D341E copy

Bitcoin

3ATGMxNzCUFzxpMCHL5sWSt4DVtS8UqXpi copy

Thank you for your support!

Follow Us!

:(