Use this function to pivot tables. This is performed by designating three columns: index, column, and value. Index is the column where pivoted rows should be aligned on; column represents the column to pivot; and value specifies the values of the pivoted table. For cases with multiple indexes for a column label, the aggregation method is to pick the first occurrence in the data frame.
All rows of a single index value must fit on one node.
The maximum rows for a single index value and column label is
Chunk size * Chunk size.