This will consolidate the three-way data down to a two-way table.
If one of the factors has little impact, it may make sense to consolidate the data so it has two factors, rather than three.
When columns are combined, the values stay the same but there are more subcolumns. So if you choose to combine columns A and B (and C and D), you'll end up with half as many dataset columns, each with twice as many subcolumns. The Y1 subcolumn of column B becomes the Y3 subcolumn of the new column A.