Auto-G-Computation of Causal Effects on a Network

. Author manuscript; available in PMC: 2021 Aug 6.

Published in final edited form as: J Am Stat Assoc. 2020 Oct 1;116(534):833–844. doi: 10.1080/01621459.2020.1811098

Gibbs Sampler I:

for m = 0, let (L⁽⁰⁾, Y⁽⁰⁾) denote initial values;

for m = 0,…, M

let i = (m mod N) + 1;

draw

L_{i}^{(m + 1)}

from

f (L_{i} ∣ L_{- i}^{(m)})

and

Y_{i}^{(m + 1)}

) from

f (Y_{i} ∣ Y_{- i}^{(m)}, a, L^{(m)});

Let

L_{- i}^{(m + 1)} = L_{- i}^{(m)}

and

Y_{- i}^{(m + 1)} = Y_{- i}^{(m)}