Table 3. Priors on the model parameters for the SIR model with change points and weekly reporting modulation.

Parameter	Variable	Prior distribution
Change points	$t_1$	$\text{Normal}\left(2020/03/09,3\right)$
	$t_2$	$\text{Normal}\left(2020/03/16,1\right)$
	$t_3$	$\text{Normal}\left(2020/03/23,1\right)$
Change duration	$\text{Δ}{t}_i$	$\text{LogNormal}\left(\text{log}\left(3\right),0.3\right)$
Spreading rates	${\lambda }_0$	$\text{LogNormal}\left(\text{log}\left(0.4\right),0.5\right)$
	${\lambda }_1$	$\text{LogNormal}\left(\text{log}\left(0.2\right),0.5\right)$
	${\lambda }_2$	$\text{LogNormal}\left(\text{log}\left(1/8\right),0.5\right)$
	${\lambda }_3$	$\text{LogNormal}\left(\text{log}\left(1/16\right),0.5\right)$
Recovery rate	$\mu$	$\text{LogNormal}\left(\text{log}\left(1/8\right),0.2\right)$
Reporting delay	$D$	$\text{LogNormal}\left(\text{log}\left(8\right),0.2\right)$
Weekly modulation amplitude	$f_w$	$\text{Beta}\left(\text{mean}=0.7,\text{std}=0.17\right)$
Weekly modulation phase	${\text{Φ}}_w$	$\text{vonMises}\left(\text{mean}=0,\kappa =0.01\right)$ (nearly flat)
Initially infected	$I_0$	$\text{HalfCauchy}\left(100\right)$
Scale factor	$\sigma$	$\text{HalfCauchy}\left(10\right)$