Description Usage Arguments Details Value See Also Examples
Functions to print or plot objects from cfboost
. Further
functions to extract the coefficients are specified. The extract
function allows to reduce the model to an earlier boosting iteration,
as e.g., the optimal stopping iteration computed with mstop
.
1 2 3 4 5 6 7 8 9 10 11 12 13 | ## S3 method for class 'cfboost'
print(x, ...)
## S3 method for class 'cfboost'
summary(object, ...)
## S3 method for class 'cfboost'
plot(x, which = NULL, ask = TRUE && dev.interactive(),
type = "b", ylab = expression(f[partial]), add_rug = TRUE,
color.palette = c("heat.colors", "terrain.colors", "topo.colors",
"cm.colors", "rainbow", "none"), ...)
## S3 method for class 'cfboost'
coef(object, ...)
## S3 method for class 'cfboost'
object[i, ...]
|
x |
an object of class |
object |
an object of class |
which |
integer vector. Only the plots for the given base-learners are plotted. Per default, only the selected base-learners are plotted. |
ask |
logic. If true a user input is required between each plot,
see |
type |
type of plot to be drawn. See |
ylab |
A title for the y axis. |
add_rug |
logic. Determines if |
color.palette |
character. Determines how time-varying effects
of non-binary covariates should be ploted. A color palette of
"heat.colors", "terrain.colors", "topo.colors", "cm.colors",
"rainbow" is specified (as character). See |
i |
integer. Index specifying the model to extract. See example for more details. |
... |
additional arguments (not used a.t.m.) |
The function print
prints basic information about the model
specified by object
and returns the argument object
invisible. summary
is a wrapper to the print
function put gives additional information on the frequency of
selection of the base-learners (see freq.sel
).
plot
gives a simple plotting interface for the estimated
effects of the model object
.
The function coef
extracts the coefficients from the model
object
.
object[i]
extracts the model from the i-th boosting iteration.
object[i]
returns the cfboost
model-object from the i-th
iteration.
cfboost
for model fitting.
freq.sel
for the selection frequencies of base-learners.
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 | ## a simple example
#################
# simulate data #
#################
set.seed(4321)
## sample covariates first
X <- matrix(NA, nrow=400, ncol=3)
X[,1] <- runif(400, -1, 1)
X[,2] <- runif(400, -1, 1)
X[,3] <- runif(400, -1, 1)
## time-constant hazard rate
lambda <- function(time, x){
exp(0.1 * time - 0.7 * x[1] + 0.5 * x[2])
}
## specify censoring function
cens_fct <- function(time, mean_cens){
censor_time <- rexp(n = length(time), rate = 1/mean_cens)
event <- (time <= censor_time)
t_obs <- apply(cbind(time, censor_time), 1, min)
return(cbind(t_obs, event))
}
data <- rSurvTime(lambda, X, cens_fct, mean_cens = 5)
##################
# estimate model #
##################
ctrl <- boost_control( mstop = 80, risk="oobag")
weights <- c(rep(1, 300), rep(0, 100))
## non-censored observations (in-bag)
sum(data$event[weights==1])/nrow(data[weights==1,])
## fit (a simple) model
model <- cfboost(Surv(time, event) ~ bolsTime(time) + bols(x.1) + bols(x.2) + bols(x.3),
control = ctrl, data = data, weights = weights)
#########################
# processing the output #
#########################
## estimate optimal mstop
(stop.opt <- mstop(model))
model_opt <- model[stop.opt]
model
## summary for mstop = 100
summary(model)
## summary (with "optimal" mstop)
summary(model_opt)
## plot of baseline hazard
plot(model_opt, which=1)
## plot of x.1 und x.2
par(mfrow=c(1,2))
plot(model_opt, which = c(2,3), ask = FALSE)
## extract coefficients
coef(model_opt)
## almost correct estimations
|
Add the following code to your website.
For more information on customizing the embed code, read Embedding Snippets.