Description Usage Arguments Details Value Note Author(s) References See Also Examples
Use an HMM to perform search for optimal system state sequence. getE
and
replicatecolumns
are two helper functions for computing the emission probabilities
and doing some internal matrix operations. perform.hmmsearch
is usually called internally.
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 | perform.hmmsearch(phi.n, bestmodel)
perform.hmmsearch_R(phi.n, bestmodel)
perform.hmmsearch_C(phi.n, bestmodel)
perform.hmmsearch_globalest(phi.n, bestmodel)
getE(x, datx, thetaprime)
replicatecolumns(mat, replicates=4)
is_consistent(phi, gammax)
checkC(phi, gammax, ch, pa)
updateA(maxima.ind, pseudocount, M, A)
viterbi(gammaposs, datx, thetaprime, T, R, M, A)
|
phi.n |
Adjacency matrix describing the candidate network. |
datx |
The data matrix. |
bestmodel |
Model object containing the current model. |
x |
State vector. |
thetaprime |
Model parameters for gaussians. |
mat |
Matrix object. |
replicates |
How often should each column be replicated. |
phi |
An adjacency matrix. |
gammax |
A state transition matrix. |
ch |
Integer denoting a child node. |
pa |
Integer denoting a parent node. |
maxima.ind |
Vector of integer. Series of state indices describing the optimal path through the system states. |
pseudocount |
Integer. The pseudocount value used in the viterbi update of the state transition matrix. |
M |
Integer. Number of states in A. Used for normalisation of the transition probability update. |
A |
The transition probability matrix. |
gammaposs |
The system state matrix containing |
T |
Integer. The number of time points. |
R |
Integer. The number of replicates. |
perform.hmmsearch
is called whenever an HMM is to be run. Depending on the global variable IMPLEMENTATION
, it decides
upon the implementation of the HMM to use. IMPLEMENTATION
can be set by passing command line argument implementation
in the function call to ddepn
. See the documentation there for a description.
is_consistent
checks whether a state transition matrix is consistent with the structure. This means, for
instance, that a child node ch
can never be activated whenever a parent pa
is active and
connected via an inhibition edge. The function returns the number of found inconsitencies and some extra
information. It is called in implementation R_globalest
to improve the assignment of active and
passive states.
TODO
TODO
Christian Bender
TODO
TODO
1 2 3 | ##---- Should be DIRECTLY executable !! ----
##-- ==> Define data, use random,
##-- or do help(data=index) for the standard data sets.
|
Add the following code to your website.
For more information on customizing the embed code, read Embedding Snippets.