This is a small release focusing on fixing the bin indexing for categorization of numeric risk factors. The change affects only the second element of functions’ output and aligns it with summary table indexing of the bins.
Changes:
NA
only, not producing the error that numeric vector is needed (check for
sc argument in the function check.init
):!is.numeric(sc)
replaced with
ifelse(length(sc) == 1, ifelse(is.numeric(sc) | is.na(sc), FALSE, TRUE), ifelse(is.numeric(sc), FALSE, TRUE))
cond.03 <- FALSE
moved to the proper place in if
clause when checking for binary target type "3" = "stop('y is not 0/1 varibale)'",
corrected to
"3" = "stop('y is not 0/1 varibale')",
Changes:
checks.iter
function corrected:
"3" = "stop('y is not 0/1 varibale')"
- variable (thanks to
Kevin Maher for spotting and reporting the typo)mdt.bin
.Changes:
mdt.bin
- correction of min.avg.rate
condition for continuous target variable.Changes:
Changes:
sc
extended for additional element -Inf
.Changes:
mdt.bin
function - identification of upper
bound for partitioning.