The estimated conditional log likelihood from a fitted model.
# S4 method for kalmand_pomp
cond.logLik(object, ...)# S4 method for pfilterd_pomp
cond.logLik(object, ...)
# S4 method for wpfilterd_pomp
cond.logLik(object, ...)
# S4 method for bsmcd_pomp
cond.logLik(object, ...)
result of a filtering computation
ignored
The numerical value of the conditional log likelihood.
Note that some methods compute not the log likelihood itself but instead a related quantity.
To keep the code simple, the cond.logLik function is nevertheless used to extract this quantity.
When object is of class ‘bsmcd_pomp’ (i.e., the result of a bsmc2 computation), cond.logLik returns the conditional log “evidence” (see bsmc2).
The conditional likelihood is defined to be the value of the density of $$Y(t_k) | Y(t_1),\dots,Y(t_{k-1})$$ evaluated at \(Y(t_k) = y^*_k\). Here, \(Y(t_k)\) is the observable process, and \(y^*_k\) the data, at time \(t_k\).
Thus the conditional log likelihood at time \(t_k\) is $$\ell_k(\theta) = \log f[Y(t_k)=y^*_k \vert Y(t_1)=y^*_1, \dots, Y(t_{k-1})=y^*_{k-1}],$$ where \(f\) is the probability density above.
More on particle-filter based methods in pomp:
bsmc2(),
eff.sample.size(),
filter.mean(),
filter.traj(),
kalman,
mif2(),
pfilter(),
pmcmc(),
pred.mean(),
pred.var(),
saved.states(),
wpfilter()