Documentation | Installation | Quickest usage | API reference | Qiita (Japanese)

CovsirPhy is a Python package for COVID-19 (Coronavirus disease 2019) data analysis with phase-dependent SIR-derived ODE models. We can download datasets and analyse it easily. This will be a helpful tool for data-informed decision making. Please refer to "Method" part of Kaggle Notebook: COVID-19 data with SIR model to understand the methods.

• Data preparation and data visualization
• Phase setting with S-R Trend analysis
• Numerical simulation of ODE models
  • Stable: SIR, SIR-D and SIR-F model
  • Development: SIR-FV and SEWIR-F model
• Phase-dependent parameter estimation of ODE models
• Scenario analysis: Simulate the number of cases with user-defined parameter values
• (In development): Find the relationship of government response and parameter values

• Monitor the spread of COVID-19
• Keep track parameter values/reproduction number in each country/province
• Find the relationship of reproductive number and measures taken by each country

If you have ideas or need new functionalities, please join this project. Any suggestions with Github Issues are always welcomed. Please read Guideline of contribution in advance.

We have the following options to start analysis with CovsirPhy. Datasets are not included in this package, but we can prepare them with DataLoader class.

Installation and dataset preparation explains how to install and prepare datasets for all users.

Stable versions of Covsirphy are available at PyPI (The Python Package Index): covsirphy and support Python 3.6 or newer versions.

pip install covsirphy --upgrade

Development versions are in GitHub repository: CovsirPhy.

pip install "git+https://github.com/lisphilar/covid19-sir.git#egg=covsirphy"

Main datasets will be retrieved via COVID-19 Data Hub and the citation is
Guidotti, E., Ardia, D., (2020), "COVID-19 Data Hub", Journal of Open Source Software 5(51):2376, doi: 10.21105/joss.02376.

Quickest tour of CovsirPhy is here. The following codes analyze the records in Japan, but we can change the country name when creating Scenario class instance for your own analysis.

import covsirphy as cs
# Download datasets
data_loader = cs.DataLoader("input")
jhu_data = data_loader.jhu()
population_data = data_loader.population()
# Check records
snl = cs.Scenario(jhu_data, population_data, country="Japan")
snl.records()
# S-R trend analysis
snl.trend().summary()
# Parameter estimation of SIR-F model
snl.estimate(cs.SIRF)
# History of reproduction number
_ = snl.history(target="Rt")
# History of parameters
_ = snl.history_rate()
_ = snl.history(target="rho")
# Simulation for 30 days
snl.add(days=30)
_ = snl.simulate()

Further information:

• Quickest version
• Quick version
• Details: phases
• Details: theoretical datasets
• Details: policy measures
• Example codes in "example" directory of this repository
• Kaggle: COVID-19 data with SIR model

Please support this project as a developer (or a backer).

Please refer to LICENSE file.

We have no original papers the author and contributors wrote, but please cite this package as follows.

CovsirPhy Development Team (2020), CovsirPhy, Python package for COVID-19 analysis with SIR-derived ODE models, https://github.com/lisphilar/covid19-sir

If you want to use SIR-F/SIR-FV/SEWIR-F model, S-R trend analysis, phase-dependent approach to SIR-derived models, and other scientific method performed with CovsirPhy, please cite the next Kaggle notebook.

Lisphilar (2020), Kaggle notebook, COVID-19 data with SIR model, https://www.kaggle.com/lisphilar/covid-19-data-with-sir-model

Reproduction number evolution in each country:
Ilyass Tabiai and Houda Kaddioui (2020), GitHub pages, COVID19 R0 tracker, https://ilylabs.github.io/projects/COVID-trackers/