Table 2 Algorithm of g-measure indicator function, I(t)

input: discretized prevalence time series, p_1:T; dominance prevalence threshold, θ (>  0).

initialization: parameter for recoding the zero-prevalence time point, ξ = 1, parameter for recoding excess time point, σ = 0, I_1:T = 0.

for t in 1:T do

If p_t == 0, set ξ = t.

If (p_t ≥ θ & ξ > σ), I_{(ξ + 1):(t − 1)} = 1, σ = t.

end for

output: discretized indicator time series, I_1:T.