frame.logreg {LogicReg}

Constructs a data frame for one or more Logic Regression models

Description

Evaluates all components of one or more Logic Regression models fitted by a single call to logreg.

Usage

frame.logreg(fit, msz, ntr, newbin, newresp, newsep, newcens, newweight)

Arguments

fit

object of class logreg, that resulted from applying the function logreg with select = 1 (single model fit), select = 2 (multiple model fit), or select = 6 (greedy stepwise fit).

msz

if frame.logreg is executed on an object of class logreg, that resulted from applying the function logreg with select = 2 (multiple model fit) or select = 6 (greedy stepwise fit) all logic trees for all fitted models are returned. To restrict the model size and the number of trees to some models, specify msz and ntr (for select = 2) or just msz (for select = 6).

ntr

see msz.

newbin

binary predictors to evaluate the logic trees at. If newbin is omitted, the original (training) data is used.

newresp

the response. If newbin is omitted, the original (training) response is used. If newbin is specified and newresp is omitted, the resulting data frame will not have a response column.

newsep

separate (linear) predictors. If newbin is omitted, the original (training) predictors are used, even if newsep is specified.

newweight

case weights. If newbin is omitted, the original (training) weights are used. If newbin is specified and newweight is omitted, the weights are taken to be 1.

newcens

censoring indicator. For proportional hazards models and exponential survival models only. If newbin is omitted, the original (training) censoring indicators are used. If newbin is specified and newcens is omitted, the censoring indicators are taken to be 1.

Details

This function calls eval.logreg.

Value

A data frame. The first column is the response, later columns are weights, censoring indicator, separate predictors (all of which are only provided if they are relevant) and all logic trees. Column names should be transparent.

Author(s)

Ingo Ruczinski [e-mail] and Charles Kooperberg [e-mail]

References

Ruczinski I, Kooperberg C, LeBlanc ML (2003). Logic Regression, Journal of Computational and Graphical Statistics, 12, 475-511.

Ruczinski I, Kooperberg C, LeBlanc ML (2002). Logic Regression - methods and software. Proceedings of the MSRI workshop on Nonlinear Estimation and Classification (Eds: D. Denison, M. Hansen, C. Holmes, B. Mallick, B. Yu), Springer: New York, 333-344.

See Also

logreg, eval.logreg, predict.logreg, logreg.testdat

Examples

data(logreg.savefit1,logreg.savefit2,logreg.savefit6)
#
# fit a single mode
# myanneal <- logreg.anneal.control(start = -1, end = -4, iter = 25000, update = 1000)
# logreg.savefit1 <- logreg(resp = logreg.testdat[,1], bin=logreg.testdat[, 2:21], 
#               type = 2, select = 1, ntrees = 2, anneal.control = myanneal)
frame1 <- frame.logreg(logreg.savefit1)
#
# a complete sequence
# myanneal2 <- logreg.anneal.control(start = -1, end = -4, iter = 25000, update = 0)
# logreg.savefit2 <- logreg(select = 2, ntrees = c(1,2), nleaves =c(1,7), 
#               oldfit = logreg.savefit1, anneal.control = myanneal2)
frame2 <- frame.logreg(logreg.savefit2)
#
# a greedy sequence
# logreg.savefit6 <- logreg(select = 6, ntrees = 2, nleaves =c(1,12), oldfit = logreg.savefit1)
frame6  <- frame.logreg(logreg.savefit6, msz = 3:5) # restrict the size