Performance

In this merge request we mostly improved performace:

• we have add the parallel_mode. With this mode set to TRUE some loops are executed using mcapply, i.e., instead of single thread it exploits all available cores in the computer.
• remove redundant searches using the function which
• replace some which calls for vapply calls to have vectorized processing.

We also fixed some bugs:

• The last threshold was not saved properly

We also improved the documentation by adding missing files.