TODO

1. Delete hello.R when I understand why it pops-up in the first place
2. Write check_fs_config as soon as I move to smth more sophisticated than TopN
3. unify camel-case and _-style function and variable names, e.g. in train
4. in read_data change the return response_column for cases when the
response_column is not renamed
5. Update the license
6. Supplement all awful trains of if else by switch?
7. add check method (if it is in caret machine learning methods)
8. update the documentation to the check config files, so that it lists all
tested options and makes the link to the config.README file
9. in configure.R in check_optimization_config, check for the values of internal
resampling. Possibly, write a small sub-function to check the values for resampling
configuration.
10. to stability.R add multi-set stability
11. Add verbosity option to train