Bugfix: Save forecast category probabilities in the correct order

Hi @nperez and @ptrascas,

Just as Núria mentioned yesterday offline, the probabilities were not being saved correctly. What happened was the arrays for each category were being appended at the beginning of the list:

# before
        probs_fcst <- append(list(Subset(fcst_probs[[ps]],
                                         along = 'bin',
                                         indices = perc,
                                         drop = 'selected')).
                             probs_fcst)

instead of at the end:

# now
        probs_fcst <- append(probs_fcst,
                             list(Subset(fcst_probs[[ps]],
                                         along = 'bin',
                                         indices = perc,
                                         drop = 'selected')))

Then the names were defined in the opposite way, so the category names did not correspond to the data.

This is fixed now, and I verified the consistency opening the files with NcToArray(). I also checked that the percentiles file is consistent (e.g. p10 < p33 < p67 < p90).

This can be merged whenever you consider.

Cheers,

Victòria