
Remove the worst variables from a model as long as they reduce the given criterion (backward elimination).
multi_backward(data, crit = mbic, ...)
an object of class big
.
a function defining the model selection criterion. You can use
your own function or one of these: bic
, mbic
, mbic2
,
aic
, maic
, maic2
.
optional arguments to crit
.
An object of class big
.
Type browseVignettes("bigstep")
for more details.
# NOT RUN {
set.seed(1)
n <- 30
p <- 10
X <- matrix(rnorm(n * p), ncol = p)
y <- X[, 2] + 2*X[, 3] - X[, 6] + rnorm(n)
d <- prepare_data(y, X)
d %>%
fast_forward(crit = aic) %>%
multi_backward(crit = bic)
# }
Run the code above in your browser using DataLab