Fig. 2.

Comparison of curves for Lombardy region. Left: y–axis on normal scale, right: on logarithmic scale. Regression line shown for $log(\mathrm{daily}\mathrm{confirmed})-log(\mathrm{daily}\mathrm{tested})\sim time$ with $95\%$ prediction band. Slope of regression with $95\%$ confidence interval: $a_{\mathrm{f}}=-0.047(-0.051,-0.043)$, this corresponds to a half-life (in days) of 14.829(13.712, 16.144). The slope of the regression $log(\mathrm{daily}\mathrm{confirmed})\sim time$ is $a_{\mathrm{raw}}=-0.025(-0.029,-0.021)$ corresponding to a half-life (in days) of 27.656(23.961, 32.698). The slope of the regression $log(\mathrm{daily}\mathrm{tests})\sim time$ is 0.022(0.016, 0.030) corresponding to a doubling time (in days) of 31.970(23.280, 42.506). Ratio of slopes for $a_{\mathrm{f}}/a_{\mathrm{raw}}=1.865$, with corresponding half-lives’ ratio: 0.536. The slope of the regression $log(\mathrm{cumulative}\mathrm{confirmed})-log(\mathrm{cumulative}\mathrm{tested})\sim time$ is $-0.016(-0.018,-0.015)$ corresponding to a half-life (in days) of 42.525(39.386, 46.208)