You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
I ran into this strange bug for the gdb engine and calculating partial response function. If two variables have a name overlap, for example, dbh and dbh_cv, there is an issue with variable selection. Instead of only one column for the current v, two columns get returned, namely the current v but also the one with a name overlap.
mboost::predict.mboost(object = self$get_data('fit_best'),
newdata = dummy_temp, which = v,
type = type, aggregate = 'sum')
A hacky solution is to use which = which(names(dummy_temp) == v) instead of which = v. This returns an integer with the column id of the current v
The text was updated successfully, but these errors were encountered:
I ran into this strange bug for the gdb engine and calculating partial response function. If two variables have a name overlap, for example,
dbh
anddbh_cv
, there is an issue with variable selection. Instead of only one column for the currentv
, two columns get returned, namely the currentv
but also the one with a name overlap.A hacky solution is to use
which = which(names(dummy_temp) == v)
instead ofwhich = v
. This returns an integer with the column id of the currentv
The text was updated successfully, but these errors were encountered: