- Improved documentation
- Deprecated `num.trees` argument (now defaults to num.trees of the `RF` argument)
- Documented internal `getsingletree`
- Shortened `getsingletree` by consolidating separate operations

#4 Fixed the issue of writing past the end of the vector, however neglected use-cases where `candidates` was not equal to `variables`.
This patch removes setting NA in `mean.index()` entirely, and applies the colname/rowname-wise fix to the created matrix instead.
This fixes the issue where in the README example, the MAA of X7 with X2 would be set to NA, instead of with X7.
This should also keep things stable against shuffled columns, or supplying variables in a different order than candidates, and should not encounter the issue of extra NAs.

It was intended to set the index of the variable with itself to be NA.
As written, this would set the index of the variable among all variables in the data set to be NA, thus creating longer vectors when the index of that variable exceeded the length of `variables`.
Fixes bug introduced in https://github.com/StephanSeifert/SurrogateMinimalDepth/commit/4039d104f14f5f0a09f1c97135550564163febd6

It was intended to set the index of the variable with itself to be NA.
As written, this would set the index of the variable among all variables in the data set to be NA, thus creating longer vectors when the index of that variable exceeded the length of `variables`.
Fixes bug introduced in https://github.com/StephanSeifert/SurrogateMinimalDepth/commit/4039d104f14f5f0a09f1c97135550564163febd6

Now using inherits()