alpha_region

[antoine4ucsd]

Hi First, congrats again for this impressive work! Working on the usa data: I was able to extract the vectors for the alpha regions If I get it right, the alpha[,1:8] correspond to the regional intercept and the alpha[,9:16] to the mobility. what I don't get is how the regional intercept and moblity are used in the equation for Rt. Could you help clarify this point?

thank you+++

[s-mishra]

Thanks, you are mistaken on how alpha_region is structured. It's dimensions are [chains, regions, regional_features]. which means alpha_region[,1,] is the intercept and alpha_region[,2,] is the coeff for averageMobility. Now X_partial_regional is of dimensions [states,N2,regional_features], which means X_partial_regional is [states,N2,2] matirx where X_partial_regional[,,1] is the intercept and X_partial_regional[,,2] is averageMobility

PS: I have assumned you have passed formula_regional as '~ 1 + averageMobility'.

[antoine4ucsd]

Thank you for clarifying the matrix structure. My confusion came from the equation printed in the report, where I interpreted Y(t,m) and alpha_region(r(m)) both as scalar signals. Could you help clarify what the coefficient and intercept components represent?

On a related note - the first dimension of alpha_region is 4000 (chains, as you referred to it). We are interpreting this as a timespan, measured in units smaller than days. is this correct?

thank you so much,

[s-mishra]

that is incorrect, it is the number of chains the sampler has run.