Output: Splits the
MyCol column on each instance of the letter
z, generating 4 new columns. If there are fewer than 3 instances of
z in a cell, the corresponding columns after the split are blank.
NOTE: Avoid creating datasets that are wider than 2500 1000 columns. Performance can degrade significantly on very wide datasetseven a much more narrow dataset. You should limit yourself to under 500 columns in your dataset.