Description Usage Arguments Value Examples
Infer is a generic function which carries out inference for a given model. For now only a pure Gaussian model is considered; however one has various options with which to carry out the Gaussian update in order to maximise use of resources, for example on linear combinations of the state-space rather than the whole space/
1 2 3 4 |
Graph |
object of class |
SW |
if 1, the Shermany Woodbury is used for inference over linear combinations, see vignette for details. This option cannot be set if linear combinations are not specified. |
Comb |
an |
List with fields Graph
(the original graph) and Post_GMRF
. The latter is an object of class GMRF
with mean and precision
given by the update. If a set of linear combinations is desired, then the list also contains a field Comb_results
, a list with entries mu
and cov
, the mean and covariance over the linear combinations respectively.
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 | ## Not run:
require(Matrix)
data(icesat)
data(surf_fe)
## First create observation object
icesat_obs <- Obs(df=icesat,
abs_lim = 5,
avr_method = "median",
box_size=100,
name="icesat")
## Now create GMRF defined over some FE basis
Mesh <- initFEbasis(p=surf_fe$p,
t=surf_fe$t,
M=surf_fe$M,
K=surf_fe$K)
mu <- matrix(0,nrow(Mesh),1)
Q <- sparseMatrix(i=1:nrow(surf_fe$p), j = 1:nrow(surf_fe$p), x = 1)
my_GMRF <- GMRF(mu = mu, Q = Q,name="SURF",t_axis = 0:6)
SURF <-GMRF_basis(G = my_GMRF, Basis = Mesh)
L1 <- link(SURF,icesat_obs)
e <- link_list(list(L1))
v <- block_list(list(O = icesat_obs, G = SURF))
G <- new("Graph",e=e,v=v)
G_reduced <- compress(G)
Results <- Infer(G_reduced)
## End(Not run)
|
Add the following code to your website.
For more information on customizing the embed code, read Embedding Snippets.