crondonm / trackingr

Python code for "Tracking R of COVID-19 A New Real-Time Estimation Using the Kalman Filter".

TrackingR

Python code for "Tracking R of COVID-19: A New Real-Time Estimation Using the Kalman Filter". Authors: Francisco Arroyo, Francisco Bullano, Simas Kucinskas, and Carlos Rondón-Moreno.

Suggested citation: Arroyo-Marioli F, Bullano F, Kucinskas S, Rondón-Moreno C (2021) Tracking R of COVID-19: A new real-time estimation using the Kalman filter. PLoS ONE 16(1): e0244474. https://doi.org/10.1371/journal.pone.0244474

Download Estimates (.CSV)

We try to update the database daily. However, as we add more and more datapoints, the computation takes longer. Currently, it takes about 24 hours to compute the new set of estimates. If you see a delay of three or more days, please let us know as this is probably due a to a technical issue with our GitHub repository. Instructions for the database are available in the README file included in the folder "Estimates-Database". The stable link to the dataset is: https://github.com/crondonm/TrackingR/tree/main/Estimates-Database

Currently we provide several files to download. The file "Database.csv" includes the whole set of estimates for each serial interval option. Each of the files called "Database_X" contains the estimates calculated setting the serial interval equal to X".

Replication Code

See README file in for detailed instructions.

Source of the Data

The original data are collected by the John Hopkins CSSE team and are publicly available online (https://github.com/CSSEGISandData/COVID-19).

Change Log

• 26/3/2021: We have updated the estimation procedure to (i) use more informative priors; and (ii) allow for intra-week seasonality. With these changes, we get estimates of R that are more consistent with the growth rate of infections seen in each country.

Questions?

You can write an email to simas [dot] kucinskas [at] hu [dash] berlin [dot] de - or to - crondonm [at] pm [dot] me - all comments and suggestions are most welcome.

Install Requirements

Before installing, to use the most recent version of pip (21.3.1 at the time of this writing). To upgrade pip on a Linux system, run

sudo -H pip install --upgrade pip

You could use a virtual environment, install the dependencies to your user or make a system-wide installation. To install all required dependencies to your current user, execute

pip install -r requirements.txt --upgrade --user

If you want to install the dependencies system-wide, run

sudo -H pip install -r requirements.txt

instead.

trackingr's Issues

data has not been updated for a week

Hi! Thanks for your work, it helps us a lot at https://github.com/owid/covid-19-data.

Based on your guideline

If you see a delay of three or more days, please let us know as this is probably due a to a technical issue with our GitHub repository. Instructions

I would like to bring to your attention that the last database update took place on January 5th 2022 (>7 days).

Hopefully, this is an easy fix :)

Data not updated since October 29th

Hi everyone.

I just noticed that the data is not updating since October 29th.

Estimates for Hong Kong

Could you add estimates for HK? Our World With Data is missing HK estimates and have replied to me that if they were available from your repo they could use them. Thanks.

Data not updated since July 15th

Hi team!
I just realised that the data has not been updated since the 15th of July.

Thanks!

US reproduction rate

Hello,

I found that the reproduction rate has not been updated for 10 days. May I ask if you will resume updating it?

https://github.com/crondonm/TrackingR/blob/main/Estimates-Database/database.csv

Thanks!

The data has not been updated since June 5

Hi, thank you for keeping the project running.

I warn you that the data has not been updated since June 5.

According to the guide, if it takes three or more days to update, should be notified.

Regards
David

Empty File under Replication Codes/

Each file under Replication Codes/ folder is empty. Is that correct?

Lag in the data (1st of December)

Hi team!
I just realised that the data has not been updated since the 1st of December.

Has there been a change in the update frequency or a problem in the automation scripts? Let me know if you need help.

This is the file that we use: https://github.com/crondonm/TrackingR/blob/main/Estimates-Database/database_7.csv

Thanks!

The data has not been updated since April 10

Hi, thank you for keeping the project running.

I warn you that the data has not been updated since April 10.

According to the guide, if it takes three or more days to update, should be notified.

Regards
David

Allow independent R calculation in countries reported by CSSE as Province/State

Currently many independent countries or geographically distant from mainland are calculated inside the main country data.

This same problem Our World in Data was having with respect to cases and death data.
They recently began to make these data independent so that they can be consulted, and that they do not affect the statistics of the main country.
https://github.com/owid/covid-19-data/blob/master/scripts/src/cowidev/jhu/__main__.py#L79

I made a program that allows you to see which countries are in this situation and can become independent of the calculation.
Otherwise you can align to the list that I shared from Our world in Data.
Our World in Data use TrackingR project, If you also separate these countries they will be able to show the R data of these countries.

Also in the example program I made them a routine to know which countries reported by CSSE are not really countries, I noticed that you have a filter but being static and as CSSE was incorporating other cases you would not be filtering them.
https://gist.github.com/3dgiordano/b2aea79f182e8c5aa59452366a90ddbc

* CSSE Sub-region to Region list
  American Samoa
  Anguilla
  Aruba
  Bermuda
  Bonaire, Sint Eustatius and Saba
  British Virgin Islands
  Cayman Islands
  Cook Islands
  Curacao
  Falkland Islands (Malvinas)
  Faroe Islands
  French Guiana
  French Polynesia
  Gibraltar
  Greenland
  Guadeloupe
  Guam
  Hong Kong
  Isle of Man
  Macau
  Martinique
  Mayotte
  Montserrat
  New Caledonia
  Northern Mariana Islands
  Reunion
  Saint Barthelemy
  Saint Helena, Ascension and Tristan da Cunha
  Saint Pierre and Miquelon
  Sint Maarten
  St Martin
  Turks and Caicos Islands
  Virgin Islands
  Wallis and Futuna

* CSSE Countries that are not countries
  Diamond Princess
  MS Zaandam
  Summer Olympics 2020

Serial interval for Omicron

There are lots of discussions that the serial interval for Omicron variant might be shorter. E.g. https://www.medrxiv.org/content/10.1101/2021.12.25.21268301v1

Would it be possible to have also estimates with shorter interval in the database onwards?