agaussian.RdAggregate Gaussians observed with the same mean and precision,
into an equivalent triplet, for use with family="agaussian"
inla.agaussian(y, s = NULL)Repeated observations. If y is a matrix, then each row represents
repeated observations. if y is a list, then each element of the list is
a vector of repeated observations. If y is a vector, then the whole vector
represents repeated observations. The optional scaling s, must have
the same format as y, ie matrix or vector.
NA's in y (and s) are removed and not used or counted.
If s is given, then the NA-pattern in y and s must be the same.
Optional fixed scaling of the precisions. Must be in the same format as y, and
have the same NA-pattern. See the documentation for details.
The output is a inla.mdata-object ready for use
with family="agaussian". See the example in the documentation.
A = matrix(1:25,5,5)
inla.agaussian(A)
#> inla.cols = 1 5
#> Y1 Y2 Y3 Y4 Y5
#> 1 50 0 5 5 11
#> 2 50 0 5 5 12
#> 3 50 0 5 5 13
#> 4 50 0 5 5 14
#> 5 50 0 5 5 15
A[1,-1] = NA
A[2,-(2:3)] = NA
inla.agaussian(A)
#> inla.cols = 1 5
#> Y1 Y2 Y3 Y4 Y5
#> 1 0.00 0 1 1 1.0
#> 2 6.25 0 2 2 9.5
#> 3 50.00 0 5 5 13.0
#> 4 50.00 0 5 5 14.0
#> 5 50.00 0 5 5 15.0