glmtree: Generalized Linear Model Trees

Description

Model-based recursive partitioning based on generalized linear models.

Usage

glmtree(formula, data, subset, na.action, weights, offset, cluster,
  family = gaussian, epsilon = 1e-8, maxit = 25, …)

Arguments

formula

symbolic description of the model (of type y ~ z1 + … + zl or y ~ x1 + … + xk | z1 + … + zl; for details see below).

data, subset, na.action

arguments controlling formula processing via model.frame.

weights

optional numeric vector of weights. By default these are treated as case weights but the default can be changed in mob_control.

offset

optional numeric vector with an a priori known component to be included in the model y ~ x1 + … + xk (i.e., only when x variables are specified).

cluster

optional vector (typically numeric or factor) with a cluster ID to be employed for clustered covariances in the parameter stability tests.

family

specification of a family for glm.

epsilon, maxit

control parameters passed to glm.control.

…

optional control parameters passed to mob_control.

Value

An object of class glmtree inheriting from modelparty. The info element of the overall party and the individual nodes contain various informations about the models.

Details

Convenience interface for fitting MOBs (model-based recursive partitions) via the mob function. glmtree internally sets up a model fit function for mob, using glm.fit. Then mob is called using the negative log-likelihood as the objective function.

Compared to calling mob by hand, the implementation tries to avoid unnecessary computations while growing the tree. Also, it provides a more elaborate plotting function.

References

Zeileis A, Hothorn T, Hornik K (2008). Model-Based Recursive Partitioning. Journal of Computational and Graphical Statistics, 17(2), 492--514.

Examples

Run this code

# NOT RUN {
if(require("mlbench")) {

## Pima Indians diabetes data
data("PimaIndiansDiabetes", package = "mlbench")

## recursive partitioning of a logistic regression model
pid_tree2 <- glmtree(diabetes ~ glucose | pregnant +
  pressure + triceps + insulin + mass + pedigree + age,
  data = PimaIndiansDiabetes, family = binomial)

## printing whole tree or individual nodes
print(pid_tree2)
print(pid_tree2, node = 1)

## visualization
plot(pid_tree2)
plot(pid_tree2, tp_args = list(cdplot = TRUE))
plot(pid_tree2, terminal_panel = NULL)

## estimated parameters
coef(pid_tree2)
coef(pid_tree2, node = 5)
summary(pid_tree2, node = 5)

## deviance, log-likelihood and information criteria
deviance(pid_tree2)
logLik(pid_tree2)
AIC(pid_tree2)
BIC(pid_tree2)

## different types of predictions
pid <- head(PimaIndiansDiabetes)
predict(pid_tree2, newdata = pid, type = "node")
predict(pid_tree2, newdata = pid, type = "response")
predict(pid_tree2, newdata = pid, type = "link")

}
# }

Run the code above in your browser using DataLab

Data engineering and BI courses are free!