discrepancy in tree files on resume from checkpoint
When resuming from a checkpoint, some of the data in the tree files differ between the original tree file and in new one.
This data is at the individual level (nb_genes_activ, nb_genes_inhib, nb_non_fun_genes, nb_coding_RNAs, nb_non_coding_RNAs)
It seems that the data being written is generally wrong (contain only zeroes) but that the data for one individual (the best one?) on resume is good. I suppose the best indiv is re-evaluated ion resume which leads to the correct values being issued while the general line of action is simply incorrect. I further suppose that this is an aevol-7 issue