Biology Reference
In-Depth Information
Then we can call the
ARTIVAnet
function, specifying the target variables with the
targetData
argument and parent variables with the
parentData
argument.
> DBN = ARTIVAnet(
+ targetData = simulatedProfiles[targets, ],
+ parentData = simulatedProfiles[parents, ],
+ targetNames = targets,
+ parentNames = parents,
+ niter = 50000,
+ savePictures = FALSE)
The number of iterations performed by the algorithm is set with the
niter
argu-
ment.
The return value of
ARTIVAnet
is a data frame containing, for each pair of
parent and target variables and for each phase, the estimated regression coefficient
and its posterior probability.
> head(ARTIVAtest1[, -7])
Parent Target CPstart CPend PostProb CoeffMean
1 TF1 1 2 10 0.0469 0.00000
2 TF2 1 2 10 0.0157 0.00000
3 TF3 1 2 10 0.0349 0.00000
4 TF4 1 2 10 0.0317 0.00000
5 TF5 1 2 10 0.0206 0.00000
6 TF1 1 11 30 0.9996 -1.52802
When
savePictures = FALSE
,
ARTIVAnet
produces several sets of plots
detailing the progress of the (RJ-)MCMC simulation. An example is shown in
Fig.
3.8
. If, on the other hand,
savePictures
is set to
FALSE
, the same plots
are saved in an output file (by default, a PDF file in an ad hoc subdirectory named
ARTIVAnet
).
Exercises
3.1.
Consider the
Canada
data set from the
vars
package, which we analyzed in
Sect.
3.5.1
.
(a) Load the data set from the
vars
package and investigate its properties using the
exploratory analysis techniques covered in Chap.
1
.
(b) Estimate a VAR(1) process for this data set.
(c) Build the auto-regressive matrix
A
and the constant matrix
B
defining the
VAR(1) model.
(d) Compare the results with the LASSO matrix when estimating the
L
1
-penalty
with cross-validation.
(e) What can you conclude?
Search WWH ::
Custom Search