Predicting New Data -

Sorry, but your current information doesn't help to solve your problem. Not
having explicit base-learners is not the cause of your problem, especially
as for glmboost() there are no explicit base-learners. You should definitely
provide a minimal example that helps to reproduce your error/problem. 

For me, everything works as intended:

## load mboost
library(mboost)

## set up model (in analogy to ?glmboost)
cars.gb <- glmboost(dist ~ speed, data = cars,

+                     control = boost_control(mstop = 400))

## without new data
p1 <- predict(cars.gb)
head(p1)

[,1]
[1,] -1.849460
[2,] -1.849460
[3,]  9.947766
[4,]  9.947766
[5,] 13.880175
[6,] 17.812584

## now with new data
p2 <- predict(cars.gb, newdata =  cars)
head(p2)

[,1]
[1,] -1.849460
[2,] -1.849460
[3,]  9.947766
[4,]  9.947766
[5,] 13.880175
[6,] 17.812584

## now with really new data
data <- data.frame(speed = c(1:10))
p3 <- predict(cars.gb, newdata =  data)
p3

[,1]
 [1,] -13.646686
 [2,]  -9.714277
 [3,]  -5.781869
 [4,]  -1.849460
 [5,]   2.082949
 [6,]   6.015358
 [7,]   9.947766
 [8,]  13.880175
 [9,]  17.812584
[10,]  21.744993


rydood wrote

Uwe Ligges-3 wrote

Hard to debug given we neither have the formula nor the data.

All I can give is that the data is in the form y, x1, x2, x3,... and the
formula is y ~ x1+x2+x3+...
Maybe it's because I'm not explicitly specifying the base learners.

--
View this message in context: http://r.789695.n4.nabble.com/Predicting-New-Data-Error-mboost-tp4654677p4654840.html
Sent from the R help mailing list archive at Nabble.com.