Enable post-pruning for covlmc models

[fabrice-rossi]

The cutoff and prune functions for vlmc enable an efficient search for an optimal model according to e.g. the BIC. A similar functionality for covlmc models would be very useful, especially considering that the estimation time for those models is significantly longer than for vlmc models.

The ground work for cutoff has already been implemented in commit a9600eb.

Post pruning needs the data to be saved in the covlmc but the added memory consumption should be compensated by the reduced total processing time when we look for an optimal model. Models fitted to the data but discarded could also be saved to speed up the process even more.

[fabrice-rossi]

The series of commits referencing this issue implement a first working version of the feature. Storing the models discarded during the first run of the algorithm (in covlmc) could speed up the pruning process, but this is non obvious as any modification in the deeper part of the tree can impact all the models above.

[fabrice-rossi]

Optimisation postponed to feature #15.