For each row, the function identifies the column that has the maximum
value. The function returns a data frame with two columns: the first is
the column name corresponding to the column of maximum value and the
second is the correspond maximum. The first column is convereted to a
factor.
If the maximum is zero, the maximum column is identified as "zero".
If there are over nbig
factors in column 1, the maximum values
that are less than the largest are combined and identified as "other".
Intended use is to transform community ecology data for use in yai
where method is randomForest.