plot.foehnix | R Documentation |
Visual representation foehnix
model optimization.
## S3 method for class 'foehnix'
plot(x, which = NULL, log = TRUE, ..., ask = TRUE)
x |
a |
which |
|
log |
logical, if |
... |
additional arguments, unused. |
ask |
boolean, default is |
There are currently three different plot types.
"loglik"
shows the log-likelihood sum path trough
the iterations of the EM algorithm for parameter estimation.
"loglikcontribution"
shows the log-likelihood
contribution (initial value subtracted; all paths start
with 0
).
coef
shows the development of the (standardized)
coefficients during EM optimization. Parameters of the
components are shown on the real scale, the coefficients
of the concomitant model (if used) are shown on the
standardized scale.
hist
plots empirical histograms of the main variable
for the two clusters (split at posterior probability >= 0.5).
In addition, the estimated parametric clusters are shown.
posterior
creates a histogram of the posterior probabilities.
point masses at 0.0
and 1.0
indicate sharp separation
of the clusters (posterior probabilities close to 0
or 1
).
The two additional input arguments breaks
(numeric vector)
and freq
(logical) will be forwarded to hist(...)
.
Reto Stauffer
Add the following code to your website.
For more information on customizing the embed code, read Embedding Snippets.