>>9 SEIRモデルの数値解なら別ステに書いたけど、結局のところパラメータが憶測に過ぎないんだよなぁ。
"
SEIR MODEL
dS(t)/dt = mu*(N-S) - b*S(t)*I(t)/N - nu*S(t)
dE(t)/dt = b*S(t)I(t)/N - (mu+sig)*E(t)
dI(t)/dt = sig*E(t) - (mu+g)*I(t)
dR(t)/dt = g*I(t) - mu*R + nu*S(t)
mu:自然死亡率 b:感染率(S->I)
nu:ワクチン有効率(S->R) sig:発症率(E->I),g:回復率(I->R)
"
SEIRモデルのパラメータ
SEIR2 <- function(
# Parameters
contact_rate = 10, # number of contacts per day
transmission_probability = 0.01, # transmission probability
beta = contact_rate * transmission_probability, # tranmission rate
infectious_period = 20, # infectious period
gamma = 1 / infectious_period, # Prob[infected -> recovered]
latent_period = 5, # latent perior
sigma = 1/latent_period, # The rate at which an exposed person becomes infective
mu = 0, # The natural mortality rate
nu = 0 , # vaccination moves people from susceptible to resistant directly, without becoming exposed or infected.
Ro = beta/gamma, # Ro - Reproductive number.
# Initial values for sub-populations.
s = 99, # susceptible hosts
e = 0, # exposed hosts
i = 1, # infectious hosts
r = 0, # recovered hosts
# Compute total population.
N = s + i + r + e,
# Output timepoints.
timepoints = seq (0, 365, by=0.5),
...
)
有病率を１％とすると、3000人にクルーズ船でも100人の屋形船でも感染者のピークは変わらないな。
同一時間あたりのcontact_rateとtransmission_probabilityが宴会での方が高いからだろうな。
パラメータを変えてグラフを書いてみた。
