Control Function for vglm()
Algorithmic constants and parameters for running vglm are set
using this function.
vglm.control(checkwz = TRUE, Check.rank = TRUE, Check.cm.rank = TRUE,
criterion = names(.min.criterion.VGAM),
epsilon = 1e-07, half.stepsizing = TRUE,
maxit = 30, noWarning = FALSE,
stepsize = 1, save.weights = FALSE,
trace = FALSE, wzepsilon = .Machine$double.eps^0.75,
xij = NULL, ...)checkwz |
logical indicating whether the diagonal elements
of the working weight matrices should be checked
whether they are sufficiently positive, i.e., greater
than |
Check.rank |
logical indicating whether the rank of the VLM matrix should be checked. If this is not of full column rank then the results are not to be trusted. The default is to give an error message if the VLM matrix is not of full column rank. |
Check.cm.rank |
logical indicating whether the rank of each constraint matrix should be checked. If this is not of full column rank then an error will occur. Under no circumstances should any constraint matrix have a rank less than the number of columns. |
criterion |
character variable describing what criterion is to be
used to test for convergence. The possibilities are
listed in |
epsilon |
positive convergence tolerance epsilon. Roughly speaking,
the Newton-Raphson/Fisher-scoring iterations are assumed
to have converged when two successive |
half.stepsizing |
logical indicating if half-stepsizing is allowed. For
example, in maximizing a log-likelihood, if the next
iteration has a log-likelihood that is less than
the current value of the log-likelihood, then a half
step will be taken. If the log-likelihood is still
less than at the current position, a quarter-step
will be taken etc. Eventually a step will be taken
so that an improvement is made to the convergence
criterion. |
maxit |
maximum number of (usually Fisher-scoring) iterations allowed. Sometimes Newton-Raphson is used. |
noWarning |
logical indicating whether to suppress a warning if
convergence is not obtained within |
stepsize |
usual step size to be taken between each Newton-Raphson/Fisher-scoring iteration. It should be a value between 0 and 1, where a value of unity corresponds to an ordinary step. A value of 0.5 means half-steps are taken. Setting a value near zero will cause convergence to be generally slow but may help increase the chances of successful convergence for some family functions. |
save.weights |
logical indicating whether the |
trace |
logical indicating if output should be produced for each
iteration. Setting |
wzepsilon |
small positive number used to test whether the diagonals of the working weight matrices are sufficiently positive. |
xij |
A formula or a list of formulas.
Each formula has a RHS giving M terms making up a
covariate-dependent term (whose name is the response).
That is, it creates a variable
that takes on different values for each linear/additive predictor,
e.g., the ocular pressure of each eye.
The M terms must be unique;
use The function |
... |
other parameters that may be picked up from control functions that are specific to the VGAM family function. |
Most of the control parameters are used within
vglm.fit and you will have to look at that to
understand the full details.
Setting save.weights = FALSE is useful for some
models because the weights slot of the object
is the largest and so less memory is used to store the
object. However, for some VGAM family function,
it is necessary to set save.weights = TRUE because
the weights slot cannot be reconstructed later.
A list with components matching the input names. A little error
checking is done, but not much.
The list is assigned to the control slot of
vglm objects.
For some applications the default convergence criterion should
be tightened.
Setting something like criterion = "coef", epsilon = 1e-09
is one way to achieve this, and also add
trace = TRUE to monitor the convergence.
Setting maxit to some higher number is usually not
needed, and needing to do so suggests something is wrong, e.g.,
an ill-conditioned model, over-fitting or under-fitting.
Reiterating from above,
setting trace = TRUE is recommended in general.
In Example 2 below there are two covariates that have
linear/additive predictor specific values. These are
handled using the xij argument.
Thomas W. Yee
Yee, T. W. and Hastie, T. J. (2003). Reduced-rank vector generalized linear models. Statistical Modelling, 3, 15–41.
# Example 1.
pneumo <- transform(pneumo, let = log(exposure.time))
vglm(cbind(normal, mild, severe) ~ let, multinomial, data = pneumo,
crit = "coef", step = 0.5, trace = TRUE, epsil = 1e-8, maxit = 40)
# Example 2. The use of the xij argument (simple case).
ymat <- rdiric(n <- 1000, shape = rep(exp(2), len = 4))
mydat <- data.frame(x1 = runif(n), x2 = runif(n), x3 = runif(n),
x4 = runif(n),
z1 = runif(n), z2 = runif(n), z3 = runif(n),
z4 = runif(n))
mydat <- transform(mydat, X = x1, Z = z1)
mydat <- round(mydat, digits = 2)
fit2 <- vglm(ymat ~ X + Z,
dirichlet(parallel = TRUE), data = mydat, trace = TRUE,
xij = list(Z ~ z1 + z2 + z3 + z4,
X ~ x1 + x2 + x3 + x4),
form2 = ~ Z + z1 + z2 + z3 + z4 +
X + x1 + x2 + x3 + x4)
head(model.matrix(fit2, type = "lm")) # LM model matrix
head(model.matrix(fit2, type = "vlm")) # Big VLM model matrix
coef(fit2)
coef(fit2, matrix = TRUE)
max(abs(predict(fit2)-predict(fit2, new = mydat))) # Predicts correctly
summary(fit2)
## Not run:
# plotvgam(fit2, se = TRUE, xlab = "x1", which.term = 1) # Bug!
# plotvgam(fit2, se = TRUE, xlab = "z1", which.term = 2) # Bug!
plotvgam(fit2, xlab = "x1") # Correct
plotvgam(fit2, xlab = "z1") # Correct
## End(Not run)
# Example 3. The use of the xij argument (complex case).
set.seed(123)
coalminers <- transform(coalminers,
Age = (age - 42) / 5,
dum1 = round(runif(nrow(coalminers)), digits = 2),
dum2 = round(runif(nrow(coalminers)), digits = 2),
dum3 = round(runif(nrow(coalminers)), digits = 2),
dumm = round(runif(nrow(coalminers)), digits = 2))
BS <- function(x, ..., df = 3)
sm.bs(c(x,...), df = df)[1:length(x),,drop = FALSE]
NS <- function(x, ..., df = 3)
sm.ns(c(x,...), df = df)[1:length(x),,drop = FALSE]
# Equivalently...
BS <- function(x, ..., df = 3)
head(sm.bs(c(x,...), df = df), length(x), drop = FALSE)
NS <- function(x, ..., df = 3)
head(sm.ns(c(x,...), df = df), length(x), drop = FALSE)
fit3 <- vglm(cbind(nBnW,nBW,BnW,BW) ~ Age + NS(dum1, dum2),
fam = binom2.or(exchangeable = TRUE, zero = 3),
xij = list(NS(dum1, dum2) ~ NS(dum1, dum2) +
NS(dum2, dum1) +
fill(NS( dum1))),
form2 = ~ NS(dum1, dum2) + NS(dum2, dum1) + fill(NS(dum1)) +
dum1 + dum2 + dum3 + Age + age + dumm,
data = coalminers, trace = TRUE)
head(model.matrix(fit3, type = "lm")) # LM model matrix
head(model.matrix(fit3, type = "vlm")) # Big VLM model matrix
coef(fit3)
coef(fit3, matrix = TRUE)
## Not run:
plotvgam(fit3, se = TRUE, lcol = "red", scol = "blue", xlab = "dum1")
## End(Not run)Please choose more modern alternatives, such as Google Chrome or Mozilla Firefox.