This set of controls are used to guide all algorithms implemented in this package.
A list variable contains several parameters for computing.
For example, exp.min and exp.max will control the range of
densities function before taking logarithm. If the density values were no
in the range, they would be rescaled. The scaling factor will be also
recorded for post adjustment for observed data log likelihood.
This will provide more accurate posterior probabilities and observed data
log likelihood.
Also, U.min and U.max will control the output of
chol when decomposing SIGMA in every
E-steps. If the diagonal terms were out of the range, a PARAM$U.check
would be set to FALSE. Only the components with TRUE
U.check will estimate and update the dispersions in M-steps
for the rest of iterations.
These problems may cause wrong posteriors and log likelihood due to
the degenerate and inflated components. Usually, this is a sign of
overestimate the number of components K, or the initialization
do not provide good estimations for parameters.
See e.step for more information about computing.
.PMC.CT stores all default controls for pmclust and
pkmeans including
algorithm |
algorithms implemented |
algorithm.gbd |
algorithms implemented for gbd/spmd |
algorithm.dmat |
algorithms implemented for ddmatrix |
method.own.X |
how X is distributed |
CONTROL |
a CONTROL list as in next |
The elements of CONTROL or .pmclustEnv$CONTROL are
max.iter |
maximum number of iterations (1000) |
abs.err |
absolute error for convergence (1e-4) |
rel.err |
relative error for convergence (1e-6) |
debug |
debugging flag (0) |
RndEM.iter |
number of RndEM iterations (10) |
exp.min |
minimum exponent (log(.Machine$double.xmin)) |
exp.max |
maximum exponent (log(.Machine$double.xmax)) |
U.min |
minimum of diagonal of chol |
U.max |
maximum of diagonal of chol |
stop.at.fail |
stop iterations when fails such as NaN |
These elements govern the computing including number of iterations, convergent criteria, ill conditions, and numerical issues. Some of them are machine dependent.
Currently, the algorithm could be
em, aecm, apecm, apecma, and kmeans
for GBD and ddmatrix (with *.dmat).
The method.own.X could be gbdr, common,
single, and ddmatrix.
Programming with Big Data in R Website: http://r-pbd.org/
# NOT RUN {
# Use set.global() to generate one of this.
# X.spmd should be pre-specified before calling set.global().
# }
Run the code above in your browser using DataLab