April 2020 / October 2023
English version / Version française
PyCoA
(Python Covid Analysis) is a Python™ framework which provides:
Time serie (cumulative) | Time series (G20) |
---|---|
MAP (OECD) | Histogram (World |
---|---|
PIE (EU) | Histogram by value (Asia) |
---|---|
Spiral plot (USA) | Yearly plot (France) |
---|---|
It is designed to be accessible to non-specialists: teenagers learning Python™, students, science journalists, even scientists who are not familiar in data access methods. A simple analysis can be performed out of the box, as well as a more complex analysis for people familiar with Python™ programming. As an example, after installing pycoa to your framework, the following few lines of code produce the four figures introducing this short documentation.
import coa.front as pycoa
pycoa.setwhom('jhu')
pycoa.plot(option='sumall') # default is 'deaths', for all countries
pycoa.plot(where='g20') # managing region
pycoa.map(where='oecd',what='daily',when='01/05/2023',which='tot_confirmed')
pycoa.setwhom('owid') # changing database to OWID
pycoa.hist(which='total_vaccinations') # default is for all countries
pycoa.hist(which='cur_icu_patients',typeofhist='pie',where='european union')
pycoa.hist(which='total_people_fully_vaccinated_per_hundred',typeofhist='byvalue',where='asia')
pycoa.plot(where='usa',which='total_people_fully_vaccinated',what='weekly',typeofplot='spiral')
pycoa.setwhom('insee')
pycoa.plot(typeofplot='yearly', what='daily', when="01/01/2019:31/12/2022", option=['smooth7','sumall'], title='Deces quotidiens totaux en France')
Since the v2.0
version, PyCoA manages also local data :
Then we get plots like the ones just below. Other databases has been added for Italy or India.
SPF data | JHU-USA data |
---|---|
cf.setwhom('spf') # Santé Publique France database
cf.map(which='tot_vacc',tile='esri') # Vaccinations, map view optional tile
cf.setwhom('jhu-usa') # JHU USA database
cf.map(visu='folium') # deaths, map view with folium visualization output
PyCoA
works currently inside Jupyter
notebook, over a local install or on online platforms such as Google Colab.
A basic demo code is available as a notebook on GitHub, on Google Colab, or on Jupyter NbViewer. Other notebooks are provided in our coabook page.
Full documentation is on the Wiki.