Splitting the dataset definition to build and apply the DM algo separately

[alainamstutz]

Hi Venexia,
Meanwhile, I worked on splitting my dataset definition into two bits, based on our discussion before the holidays:

1. generate_dataset_dm_algo (extract the min. dataset needed for the DM algo) and process it and run the DM algo (data_process_dm_algo), resulting in a dataset with just 8 variables
2. generate_dataset (extract remaining variables) and process it and merge it with the output from data_process_dm_algo (data_process)

Now, I can get a larger dummy dataset (10'000 instead of 5000).

I used the base merge function (data_processed <- merge(data_processed, data_processed_dm_algo, by = "patient_id", all.x = TRUE), see data_process) hope that's ok.

Otherwise, I have not changed anything. If you have some time to double-check it (esp. the yaml), I would appreciate!
Thank you very much,
-Alain