Boosting algorithms for fitting generalized linear, additive and interaction models to potentially high-dimensional data. The current relase version can be found on CRAN (http://cran.r-project.org/package=mboost).
I find that that when mboost subsets to remove missing values here it fails to subset the weights so the dimensions don't line up down the road.
Example:
library(mboost)
# generate random data
set.seed(2017-5-22)
weights <- sample(1:100, 100, replace=FALSE)
x <- rnorm(100)
y <- runif(100)
# create missing value
x[25] <- NA
myData <- data.frame(x=x, y=y)
# errors
mboost(
y ~ bols(x),
data = myData,
weights = weights,
family = Gaussian()
)
# works
mboost(
y ~ bols(x),
data = myData,
weights = weights[-25],
family = Gaussian()
)
# base R modeling functions subset weights
lm(y ~ x, myData, weights = weights)
Hello,
I find that that when
mboost
subsets to remove missing values here it fails to subset the weights so the dimensions don't line up down the road.Example:
Kind Regards, Carl Ganz