Python package to analyze covid propagation using JHU, ECDC or Italian protezione civile (DPC) databases.
It provides a small library of functions to parse and analyze data using simple numpy arrays.
This is a ranking of regions based on a simple Principal Component Analysis (PCA) analysis of 4 parameters, all intensive in time and space (calculated per inhabitant and over the last two weeks)
- test positivity
- cases variation
- deaths variation
- cases in ICU variation
The PCA gives one main eigenvector, which accounts for almost 90% of the fluctuations, to which all parameters contribute nearly equally and with equal sign. This is tentatively taken as an alarm level indicator. It does not account for the resilience of the local health system.
Daily increase (lines are 1 week averages)
Daily new cases per 1000 inhabitants in Italian regions (1 week average)
Daily deaths per million inhabitants in Italian regions (1 week average)
Time series of the daily new cases per 1000 inhabitants in Italian regions, color coding is by latitude (not averaged)
Daily new cases in North-East cities (2 week average)
Daily new cases in European countries (>10M population). The daily increase is averaged over a sliding window of 14 days.
Daily new deaths in European countries (>10M population). The daily increase is computed from the average increase over 14 days.
Daily new cases in selected English-speaking countries. The daily increase is averaged over a sliding window of 14 days.