# work our way backwards through the time series, randomly sampling confidence leve
U = [0] * T_idx # a place to store this replicate
U[-1] = U_T
impossibleFlag = False
for t in tV: # work our way backwards
alpha = uniform.rvs()
S0 = S[t - 1]
S1 = S[t]
U1 = U[t]
d0 = d[t - 1]
U[t - 1], impossibleFlag = find_U0_bnd(alpha, S0, S1, U1, d0,
impossibleFlag, omega,
biasedurn)
# calculate X_t from U_t
N = S[0] + E[0] + U[0] # assumes X(0) = 0
X = [N - EE - SS - UU for EE, SS, UU in zip(E, S, U)]
# append results
UM.append(U)
XM.append(X)
# calculate statistics on sample
# ---
U = [0] * T_idx # a place to store this replicate
U[-1] = U_T
impossibleFlag = False
for t in tV:
S0 = S[t - 1]
S1 = S[t]
U1 = U[t]
d0 = d[t - 1]
# if type_of_ci == 'midp':
alpha = stats.uniform.rvs()
min_poss_U0 = max((min_poss_UV[t - 1], U1 + d0))
U[t - 1], impossibleFlag = find_U0_bnd(alpha, S0, S1, U1, d0,
impossibleFlag, None, None)
# store this example
UV.append(U)
# plot them
# ---
# plot the simulation as "true" values
plt.plot(S_orig, 'green', lw=1, label=r'$S_t$')
plt.plot(E_orig, 'red', lw=1, label=r'$E_t$')
plt.plot(X_orig, 'blue', lw=1, label=r'$X_t$')
# plot each of the examples
for i, U in enumerate(UV):