Skip to content

ResearchLuxembourg/restimator

Repository files navigation

REstimator – estimator of R(t) for COVID-19

Brief explanation of the indicator

And related ones (source: science.lu, "Coronavirus technical terms – explained by scientists from Luxembourg" )

R0

The “basic reproduction number” at the beginning of epidemic (time "zero"). It represents the average number of cases each infected person will likely cause if no action is taken and the whole population is susceptible. It is disease and variant-specific.

R_t

As the epidemic evolves and measures are taken, the “reproduction number” might vary in time. Likewise, from the very first value R0, the index evolves in time, tracking the infectious curve. In this regard, R0 is an upper value for R_t.

R_eff

The “effective reproduction number”, signifying the average number of cases each infected person will likely cause during the epidemic. It evolves in time like R_t, but is scaled according to the true number of susceptible people (while R_t assumes that 100% of the population is susceptible).

R_eff is used as an epidemic “thermometer”: R_eff<1 indicates a decreasing curve of daily infections (sub-linear increase of cumulative cases), R_eff=1 indicates a stable curve (linear increase of cumulative cases), R_eff>1 indicates a growing daily curve (exponential increase of cumulative cases). The higher R_eff, the more pronounced the exponential growth.

Input

Excel (.xlsx) file with at least the following columns, ordered chronologically (closest to farthest date):

  • report_date
  • new_cases
  • new_cases_resident

Output

  • R_eff, calculated for each day data are available (reported as .csv file and pdf plot)
  • R(t), calculated for each day data are available (reported as .csv file)

How-to

Automated way

If you place your .xlsx input into directory input, everything can be run just by executing ./run_pipeline.sh.

This requires a Docker image researchluxembourg/restimator pulled or built -- you can pull it from the github packages, or build manually from this repository.

Install dependencies

You need julia installed, preferably a version higher than 1.6.

You can run Julia in the directory of restimator, using the project environment as:

cd restimator
julia --project=.

After that, type

Pkg.instantiate()

This should install the necessary Julia packages.

Run the R estimators

We assume your input is placed in input/filename.xlsx. The input excel file must contain a single sheet with several correctly marked columns. You can check the suitability of the input file using script components/check_input.jl, as follows:

julia --project=. components/check_input.jl input/filename.xlsx

You may get an output like this (or eventually an error with a description of the problem; for basic troubleshooting, refer to the section below):

┌ Warning: Last entry is older than 1 week!
└ @ Main ~/work/restimator/components/check_input.jl:36
[ Info: Age of the last entry is 136 days
[ Info: Check finished OK.

Reff and Rt analyses may be run as follows:

julia --project=. components/estimate_r_eff.jl input/filename.xlsx
julia --project=. components/estimate_r_t.jl input/filename.xlsx

Without errors, the output is generated into corresponding files:

input/date_Reff_estimate.csv
input/date_Rt_estimate.csv

Basic troubleshooting

The pipeline might raise errors if the initial check on data quality is not satisfied. To understand common sources of errors and warnings, refer to this section.

  • File does not exist: error in loading the input file.
  • Incorrect file format: expected format is Excel .xlsx.
  • Incorrect file name: the de facto agreed file naming is "clinical_monitoring_'+DATEOFTODAY[yyyymmdd]+_cleaned_case_and_hospital_data".
  • Missing daily data: the program needs a data entry for each day, as a positive integer.
  • Typos in the input file (in particular, related to the input columns labels) or missing input column: expected input columns are "report date", "new_cases", "new_cases_resident".
  • Inconsistency: daily cases for residents should be less or equal to total new daily cases.
  • Retrospectively changed data. To provide consistent results, the program needs initial conditions: the data history should not be altered, starting from 2020-02-28.
  • Last datapoint missing (relative to the latest detection date of 'yesterday').

Credits and contacts

For basic troubleshooting of raised warnings and errors, check the readme section first. If the problem persist, please contact the developers.

Copyright 2020-2022 Luxembourg Centre for Systems Biomedicine

Licensed under the Apache License, Version 2.0 (the "License"); you may not use this file except in compliance with the License. You may obtain a copy of the License at

http://www.apache.org/licenses/LICENSE-2.0

Unless required by applicable law or agreed to in writing, software distributed under the License is distributed on an "AS IS" BASIS, WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. See the License for the specific language governing permissions and limitations under the License.