Possible to include more than one observation and different observations for different forecasts. Also improved the figure and parameters

Improve all functions development within s2dv_cube new structure and correct minor documentation format

Sorted the factor levels in the order required in the legend to avoid problems later. Reorganized ggplot calls.

Moved the fix to the right place. Modified it to be a onliner. Commented it. Added 0 instead of NA to the missing columns. Included the same fix for pct2 (extremes).

If there is a unique forecast, extreme limits and tercile limits must be vectors. Now 1-D array is cot accepted. Observations are proven to be a vector or a numeric value.

Corrected bug in labels of tercile/extreme probabilities. Now uses first panel positions for all labels.